the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
| 25 Jul 2023
A global social activation model of enabling conditions for positive social tipping – the role of sea-level rise anticipation and climate change concern
Abstract. Effective climate change mitigation necessitates swift societal transformations. Social tipping processes, where small triggers initiate qualitative systemic shifts, are potential key mechanisms instigating societal change. A necessary foundation for societal tipping processes is the creation of enabling conditions. Here we assess future sea-level rise estimates and social survey data within the framework of a social activation model to exemplify the enabling conditions for tipping processes. We find that in many countries, climate change concern is sufficient, the enabling conditions and opportunities for social activation already exist. Further, drawing upon the interrelation between climate change concern and anticipation of future sea level rise, we report three qualitative classes of tipping potential that are regionally clustered, with greatest potential for tipping in Western Pacific rim and East Asian countries. These findings propose a transformative pathway where climate change concern increases the social tipping potential, while extended anticipation time horizons can trigger the system towards an alternative trajectory of larger social activation for climate change mitigation.
- Preprint
(3768 KB) - Metadata XML
- BibTeX
- EndNote
Status: closed
-
RC1: 'Comment on egusphere-2023-1622', Anonymous Referee #1, 28 Aug 2023
General Comments
The manuscript is well-written and addresses the very important topic of better understanding enabling conditions for social tipping towards climate action. Quantitative approaches for socio-ecological systems are urgently needed to complement existing research for societal transformation towards climate resilience in its many facets. As this transformation is urgent and requires broad societal agreement, understanding the pathways for shift perception into an aware and ‘climate active’ state is timely and this approach provides a valuable contribution in this regard.
While the framing and approach are valuable, there are a number of points that would require additional work and analysis for the paper to achieve its full potential.
One of the key issues I see with the paper is that it focusses on one very important and also detrimental climate impact, but one that is very localised in that it applies to coastal regions only. While these are home to large cities and a large share of the global population, including in higher density settings, nonetheless it is for many regions not one of the most pressing impact and not as globally applicable (as the authors also point out in their manuscript). The anticipation horizon, as also discussed and assessed in the paper, is another key issue with SLR that, in my view, does not make it the ideal impact to focus on.
Another concern relates to the idea of ‘active population’ and how these active groups then lead to the transformative shift in climate action. I think more granularity and context is needed here. Concern does not necessarily translate into action and active populations don’t necessarily trigger the needed policy-shift.
Some of the more specific comments below provide further granularity to these to concerns, which hopefully help in revising the manuscript.
Specific comments
There are a number of globally relevant impacts that are already detrimental today and affect people globally. The most detrimental effects of SLR are very long-term and thus often operate on timescales beyond human comprehension (see Figure 1, for example), as the authors also point out in the para p2-3 l30-35. Those impacts of SLR felt today are often very localised to smaller coastal communities and even in highly affected areas today, awareness of cc let alone activism is not present. It is not sufficiently clear to me why SLR would be chosen as the relevant impact variable, when other impacts might be much more central in the current perception. As the authors point out on p3 l50-55, “Experienced climate impacts such as floods and heat waves have the potential to shift attitudes and behaviours towards climate change and instigate social tipping processes”. Why not focus on one of these widely experienced impacts for the model?
In the end, which impact actually triggers the enabling conditions for mitigation doesn’t matter as the result is the same across impacts. I would therefore disagree with the statement on p3 l73/74. In addition, the cited reference from 2012 seems both outdated in light of currently experienced impacts and the paper also does not seem to substantiate the claim made in the sentence.
It is not quite clear what is defined as ‘active population’ and how that relates to mitigation action. Of course, awareness of the population is an essential ingredient, but awareness is far from activism, let alone government follow-through, which would be what is needed to actually mitigate in line with the Paris Agreement. An EIB study found that for Germany, for example, 63% of the population in 2021 would favour stricter climate measures. According to my reading of “active” within the study, this would then be the majority of the German population. However, we do not see a tipping in mitigation action.
While I see the importance of the model developed in this study, I think it needs more nuancing in terms of what is can actually show and how this activation of tipping then leads to large-scale global action. Maybe this would be better placed within the specific environment of policy-makers and how tipping can be induced with them?
I would also like to get more context for the consideration of ‘certainly active’ population as defined on p.6 line 129. While I see the point of assuming these should be ‘certainly active’ in terms of climate awareness, this is not necessarily the case for various reasons (e.g. US in Florida; West Africa). There are certainly no large activist movements there, despite the obvious current impacts and future exposure.
Maybe it would be worth further specifying (and calibrating) the model around a specific type of climate engagement within a specific group to first understand the context under which these groups have been activated to then be able to extrapolate to that specific type of action. The extension by which the activated population would then be able to force policy-action would then need to be an additional (albeit essential) extension of the analysis.
As mentioned above, I think it is critical to further define what is meant by ‘active’. Large parts of the population are convinced that climate action is needed, but are not openly active about it. How would these be placed within the model? In this context, adding different degrees of ‘active’ – along the lines of ‘aware’ to ‘willing to engage in existing activism’ to ‘active’ might be useful to then understand how activism may trickle down. Simply having an ‘aware’ population (as seems to be the definition of active in this paper) does not lead to the action that would be need to address the climate challenge. The authors provide quite a nuanced discussion of this in Sec. 1.1, but I’m not sure I see this reflected in the final assessment. Linking back to my comment above, this could also be linked to the reflection around which climate impacts drive awareness and action and create the enabling conditions for change.
Finally, I think the discussion of enabling conditions (line 45 and following) might require the consideration of some additional literature that provides quite a lot of a additional granularity (see e.g. IPCC WG2 Ch17 on adaptation and risk or IPCC WG3 Ch17 on Transitions). Clearly for a modelling approach, some reduction of complexity is needed, but I feel that inclusion of some of this complexity, including interactions between different aspects of enabling conditions, would be needed. As per some of my comments below, I feel results to some extent appear obvious out of the model set-up and if this complexity is not accounted for, I’m not convinced a model is actually needed for the results we get.
Results
- 12 lines 287-289 isn’t this conclusion rather obvious from the model set-up?
- 14 lines 320/325 These are super interesting results and also point towards what I mentioned in the General Comments above: there are different types of impacts that would most likely be a good predictor of climate action. In the Pacific with above average SLR rates, SLR may well be the perceived most pressing impact (though tropical storms may well contribute to the awareness as well). It may be interesting to choose regionally specific impact drivers that better reflect regionally specific risks , as this would likely provide a much closer to reality situation.
Discussion
The discussion raises a number of important points and also highlights the importance of better understanding social tipping processes to target interventions where they are most likely to yield results. Exactly due to the importance of such work, I would strongly encourage a further sharpening of the analysis to be more directly relevant for understanding the complexities.
In summary, as outlined above, on the one hand, I think the approach would strongly benefit from considering further key impacts that drive awareness globally. Unfortunately, the recent years have given plenty of examples of what these may be (wildfires, droughts, heatwaves, flooding…). On the other hand, more nuancing of what ‘active’ means and how this translates into the needed policy action would be important to include.
Citation: https://doi.org/10.5194/egusphere-2023-1622-RC1 -
AC1: 'Reply to Reviewer 1', Keith Smith, 20 Mar 2024
Thank you very much for these comments and constructive feedback on the manuscript. We have grouped each of our responses thematically [listed in brackets] and have also linked to where we have responded to similar comments by other expert reviewers as well. Again, thank you again very much for the efforts and support of this manuscript.
[Why does the paper focus on SLR and not on other impacts? Comment #1]
One of the issues raised is that the analysis of this paper focuses on sea-level rise, as SLR is particularly localised. This concern is valid, and well-taken by the authorship team, but this is also true of most other projected future climate impacts. For example, floods, droughts, increased intensity and frequency of hurricanes likely share similar spatial impacts. This comment also raised the issue of the temporal distance between mitigating actions and the manifestation of SLR. This is particularly the motivating reason why we chose to focus this empirical modelling work on SLR, as this is an impact in which there is limited observed data. Rather, the primary question of this research is how a problem like anticipation of future SLR can (or cannot) lead towards increased climate mitigating behaviours and policies. There are many other papers which focus on the impact of extreme weather events (often using observed data after exposure to the event) - but for the future impacts of SLR on climate actions, this is something that can likely only be assessed via a modelling exercise such as this one (particularly on the global scale). Further to this point, we believe that SLR is a particularly unique potential climate impact not only because of the temporal disconnect between mitigating actions now are need to minimise impacts that will manifest in the future, but also importantly, SLR is a very unique potential impact of climate change, as it is particularly persistent. The impact of floods (for example), while potentially causing similar amounts of damage to localised communities and land, are comparatively more temporary. Indeed, we see that in many areas, communities rebuild on the same flood lands within comparatively short periods of time (years). While for SLR, the impacts persist over potentially thousands of years - which for current human societies is effectively permanent. In this case, the impact of SLR is a very risky form of climate change induced event, and one that is comparatively less understood within the social science and empirical literature. Accordingly, we suggest that the revised manuscript would further develop the argument for why SLR is a unique and important climate change induced impact, as well as further developing the existing literature on other impacts to note the differences and similarities with SLR.
[Theory of change and model specification - how do people translate into active population]
We thank the reviewer for bringing up these questions about the ‘active’ population as well. We agree that concern does not necessarily translate into action, this is similar to the problem of an ‘attitudes-behaviour’ gap. Our model intends to incorporate this uncertainty, where concern is rather viewed as a necessary - but not sufficient - condition for action. We parameterise the ‘potentially active’ population using the level of concern for a given country - which simply means that the person has the potential to become activated, but would only do so if a majority of their neighbour nodes become activated. This is clearly reducing a great amount of complexity in human behaviour - but given the model resolution, we explicitly do not intend to deterministically explain how people will become active. Still, we note that such an approach is in line with a comparatively large body of literature on models of social activation that effectively group actors into similar categories like the ones we employ here. Given that our model explores the complexities of the relationship between individual attitudes, network effects, SLR exposure and impact time, we keep the level of specificity for human behaviour as relatively low, as otherwise, the model components would be very difficult to identify and parameters difficult, if not impossible, to estimate. Thus, human behaviours would likely be at a different depth of specification and out of line with the other model components.
[Why does the paper focus on SLR and not on other impacts? Comment #2]
We thank you for the further comment regarding the choice of SLR as the chosen climate impact we assess in this model. Similar to our response above, we note that SLR is a unique, and comparatively understudied future climate impact. There is a comparatively large literature on the role of extreme weather events affecting attitudes and behaviours - yet even within this literature, the effects vary substantially by operationalization of the independent (event) and dependent (attitude and behaviour). For a recent finding in this regard, Figure 6 and 7s of this pre-print from Quoss and Rudolph (2023) nicely demonstrate substantial noise in the impact of extreme temperature, precipitation and other events shaping attitudes in Switzerland.
That is, we also recognise the difficulty in determining what will actually “cause” social tipping, and in a revised manuscript, would further note the limitations of any extreme event in initiating tipping dynamics - as there is limited empirical evidence that this has yet to occur.
At the same time, we further believe that SLR is an interesting case study, with the potential to motivate action - even though the impacts are indeed likely (hopefully) quite far into the future. For example, a related paper by the authorship group (under review) uses survey data from the United States (a locale with comparatively lower proportion of citizens likely to be affected directly by SLR - Class III), finding that concern about SLR is one of the top factors driving climate change policy support and behavioural willingness (see Smith et al., 2022).
Accordingly, this paper focuses on understanding how SLR would link current actions with impacts that will not manifest until generations into the future via the mechanism of anticipation. Indeed, this may be an optimistic - but important question - as actions in response to experienced SLR would already be far too late to substantially mitigate these quasi-permanent changes to vast sections of inhabited land around the world.
Lastly, in this paper we do propose a linkage between SLR and other climate impacts, where anticipation of SLR can provide grounds for event-induced tipping, i.e., the anticipation of SLR can raise concern sufficiently high so that other, more immediate impacts, can kick the system into a new stable state.
[Salience of SLR and how SLR perceptions shape attitudes and behaviors]
This point is very well taken regarding the claim made about the salience of SLR and the reference citation. We have further reviewed the literature on the salience and impact of sea-level rise. Several studies have noted that sea level rise is not particularly well understood by the public (in terms of scientific accuracy, e.g. Thomas et al., 2015, Priestley et. al, 2021), is seen as occurring quite far into the future (Covi et al., 2016), and some studies have also found that SLR does not receive frequent media attention (e.g. Akerloff et al., 2017; 2019). Yet, communicating the risks of SLR can increase climate change attitudes - even amongst more ‘hard to reach’ audiences such as Republicans in the United States (Bolsen et al., 2018; Smith et al., 2022). Accordingly, we suggest revising the manuscript to further include these relevant literatures, and develop a similar, but re-framed argument:
While SLR may be, at times, be misunderstood and not a priority for some members of the public, it has the potential to be a strong driver of climate change attitudes and behavioral change. Yet, the impact of anticipation of SLR on climate change attitudes and behaviours may be limited if the distance to projected impacts extends far into the future. Accordingly, we model how the likelihood of attitudes and behavioral change relates to differing levels of concern, being spread throughout networked dynamics, at varying levels of impact severity and timeframes.
[Definition of active population]
We agree with this comment and those further raised below - a similar concern to those brought by Reviewer 2 and 4 - regarding the lack of definition for the construct ‘active population’. We intended to draw upon an (intentionally simplified) categorization of the population into three groups - certainly active, potentially active, and not active. For the active population in this regard, these would be people that are already engaging to mitigate future climate impacts in their local settings. Given the cross-national design, the ‘action’ is relative to the needs and capacity for a given setting. This could be political action (e.g. voting behaviour directly for climate change policies, protesting), individual mitigating behaviours (e.g. pro-climate behaviours), or even social actions (spreading of climate communication, shifting norms). Given the resolution and cross-national implementation of this model, we cannot define more specifically what ‘action’ means, as any given action for any given context and person will a diverse set of motivations and barriers (complexity would need to be explored within a different empirical approach). Furthermore, we also note that the capacity to action for a given individual is likely inverse to the expected impacts of SLR - a distributional inequity posed by climate change. We would propose further revising the manuscript to better elaborate on the definition of climate action, and these distributional impacts.
Furthermore, we also suggest changing the terminology to be more reflective of the social science literature. The most broad phrase that is commonly used is ‘pro-climate change behaviors’. Accordingly, would suggest changing the terminology from climate action to ‘pro-climate change behaviors’ to be more reflective of these literatures.
Also, given the concerns regarding the terminology used for the modelling approach raised by Reviewer 2 [Model terminology and relation to other theories of environmental social change], we suggest revising the title to be “A global threshold model of enabling conditions for social tipping in pro-climate change behaviours – the role of sea-level rise anticipation and climate change concern”
We would further suggest that future research should be engaged to elaborate the mechanisms by which action can develop. Such work would need to be well-tailored to the individual action and setting. There is much work in the regards (largely enacted by environmental social psychologists, public policy researchers and behavioural economists) focusing on specific settings and actions - but disproportionately focusing on Western, wealthier states. Our goal of this modelling exercise was to intentionally to be rather abstract - allowing for cross-national comparisons. But these findings should be seen as in compliment to this rich literature exploring the more specific and diverse mechanisms of environmental change (individually, socially and politically).
Lastly, in regard to the EIB study regarding the desire of 63% of Germans to prefer stricter climate actions, we would rather consider this to be akin to the potentially active population, those willing to act, rather than to the actually active population. Responding positively to a survey question more likely displays a willingness to change, while the actual observed action would be whether this translates into political action (voting). And, as this comment notes, we did not observe tipping like behaviour within the German system (at least not yet). For example, there was a historic increase in support for the Green Party in the 2021 German Federal Election - where climate change was one of the largest issues across all parties. Yet, in the end, the Green Party only received ~15% of the popular vote and seats in Parliament. This resulted in the Green Party being part of the governing coalition (for the first time in 15 years), and likely contributed to many federal actions to combat climate change, but did indeed fall short of rapid, systemic changes. We would read this as evidence of a system that is getting closer towards change, but one that has, as of yet, not tipped.
[Specification of model and policy relevance of findings - Comment #1]
As noted in the previous comment, we also agree with the importance of developing specific mechanisms for how tipping can be induced, and proscribing actions for policy-makers and stakeholders to incentivise these changes. This is important future work - empirical research that would rely upon observed individual data (potentially within experimental settings) to understand how such changes can occur for a given context and towards a specific change. For example, there is another pre-print that focuses on expert elicitations of tipping dynamics towards decarbonization in the UK political system (Smith 2023). Yet, we would also note that defining such a specific mechanism is not the purpose or design of our research question and modelling exercise. This would require a far greater level of specificity, using different methods, and would almost certainly not be generalizable across such a cross-national design. We would revise our future research statement in a revised manuscript to make this call for further research more clear and direct.
Yet, in terms of political relevance from our modelling approach, we emphasise the role of creating enabling conditions for change to occur (e.g. Tabara et al., 2018; 2023). We find, that on it’s own, concern may be insufficient to induce tipping dynamics, but rather, provides the necessary conditions for such to change to take place (either induced by other events, policy changes, action etc). That it, concern pushes the system closer towards the threshold, such that tipping can occur.
[Specification of model and policy relevance of findings - Comment #2]
We further thank you for the comments regarding the specificity and complexity of this model. As we have noted above, we very much agree with the need for further research (across many disciplines) that further explores this complexity. For example, within a modelling exercise, research on interactions using a similar modelling design could also be engaged. Members of the authorship team have also engaged in such work using a similar model (e.g. Müller et al., 2021). We will further elaborate on different types of future research that could expand upon these findings (from an experimental, observational and modelling approach) within the revised manuscript.
We would also contend that, like most models, some of the conclusions are more directly related to the modelling design (that is, the more people are exposed to direct impacts of SLR the more likely they are to act). Yet, we contend that others are not as “obvious”. For example, the differing role of concern in shaping the likelihood of action. This is illuminated within the three classes of tipping section of the results - where we find that yes, within class I, the results are rather expected, but in classes II and III, there are interesting interactions between SLR and concern that are exposed, which we believe are non-trivial and not necessarily anticipated. We would suggest revising the manuscript to further emphasise the results that go beyond the “top line” findings, that may be more related to the modelling design itself.
Citation: https://doi.org/10.5194/egusphere-2023-1622-AC1 -
AC5: 'References for all responses', Keith Smith, 20 Mar 2024
As a final note, please do find our references for these responses to all reviewers here:
Akerlof, K., Covi, M., & Rohring, E. (2017). Communicating sea level rise. In Oxford Research Encyclopedia of Climate Science.
Akerlof, K., Merrill, J., Yusuf, J. E., Covi, M., & Rohring, E. (2019). Key beliefs and attitudes
for sea-level rise policy. Coastal Management, 47(4), 406-428.
Bolsen, T., Kingsland, J., & Palm, R. (2018). The impact of frames highlighting coastal flooding in the USA on climate change beliefs. Climatic Change, 147, 359-368.
Coleman, J. S. (1994). Foundations of social theory. Harvard university press.
Covi, M. P., & Kain, D. J. (2016). Sea-level rise risk communication: Public understanding, risk perception, and attitudes about information. Environmental Communication, 10(5), 612-633.
Diekmann, A., & Preisendörfer, P. (2003). Green and greenback: The behavioral effects of environmental attitudes in low-cost and high-cost situations. Rationality and Society, 15(4), 441-472.
Kollmuss, A., & Agyeman, J. (2002). Mind the gap: why do people act environmentally and what are the barriers to pro-environmental behavior?. Environmental education research, 8(3), 239-260.
Marzeion, B., & Levermann, A. (2014). Loss of cultural world heritage and currently inhabited places to sea-level rise. Environmental Research Letters, 9(3), 034001.
Milkoreit, M. (2023). Social tipping points everywhere?—Patterns and risks of overuse. Wiley Interdisciplinary Reviews: Climate Change, 14(2), e813.
Müller, P. M., Heitzig, J., Kurths, J., Lüdge, K., & Wiedermann, M. (2021). Anticipation-induced social tipping: can the environment be stabilised by social dynamics?. The European Physical Journal Special Topics, 230(16-17), 3189-3199.
Priestley, R. K., Heine, Z., & Milfont, T. L. (2021). Public understanding of climate change-related sea-level rise. PLoS One, 16(7), e0254348.
Quoß, F., & Rudolph, L. (2022). Operationalisation matters: Weather extremes as noisy natural experiment show no influence on political attitudes. https://osf.io/preprints/osf/m5rz8
Smith, E. K., Eder, C., Donges, J. F., Heitzig, J., Katsanidou, A., Wiedermann, M., & Winkelmann, R. (2022). Domino Effects in the Earth System-The role of wanted social tipping points. OSF Preprints.
Smith, S. R. (2023). Enabling a political tipping point for rapid decarbonisation in the United Kingdom. EGUsphere, 2023, 1-21.
Tàbara, J. D., Frantzeskaki, N., Hölscher, K., Pedde, S., Kok, K., Lamperti, F., ... & Berry, P. (2018). Positive tipping points in a rapidly warming world. Current Opinion in Environmental Sustainability, 31, 120-129.
Tàbara, J. D. (2023). Regenerative sustainability. A relational model of possibilities for the emergence of positive tipping points. Environmental Sociology, 9(4), 366-385.
Thomas, M., Pidgeon, N., Whitmarsh, L., & Ballinger, R. (2015). Mental models of sea-level change: A mixed methods analysis on the Severn Estuary, UK. Global Environmental Change, 33, 71-82.
Wiedermann, M., Smith, E. K., Heitzig, J., & Donges, J. F. (2020). A network-based microfoundation of Granovetter’s threshold model for social tipping. Scientific reports, 10(1), 11202.
Citation: https://doi.org/10.5194/egusphere-2023-1622-AC5 -
AC6: 'Notes on organization of responses to reviewer comments', Keith Smith, 20 Mar 2024
First of all, we thank the reviewers for this constructive and well considered review of our manuscript. We will address these issues raised by each point below. In order to best organise our responses, we have structured these by Reviewer (underlined) and thematic concept being addresses [in brackets]. If similar issues are raised across reviewers, we address these more directly once, and cross-reference of our responses where appropriate for clarity.
Again, thank you all very much for these constructive and helpful feedback.
Citation: https://doi.org/10.5194/egusphere-2023-1622-AC6
-
RC2: 'Comment on egusphere-2023-1622', Anonymous Referee #2, 19 Nov 2023
The paper considers the urgent need for a rapid societal-scale transformation or social tipping in response to the threat of climate change impacts. It tests a hypothesis based on a ‘social activation model’ that there is a correlation between climate change concern and anticipation of long-term sea level rise. It employs three categories of tipping potential and finds the greatest potential for tipping in Western Pacific Rim and East Asian countries.
The topic of social tipping points is very popular and timely in the context of the upcoming UNFCCC COP 28 Conference and ongoing negotiations to limit global warming. The idea to combine estimate of future seal level rises with social survey data within a model of social activation is a novel methodological approach. The use of the MAGICC climate model, and the use of topographic and population distribution data, deserves credit. The paper is well structured and composed, with appropriate number and quality of references and strong supplementary materials, clear illustrations and correctly defined and used mathematical symbols.
Specific comments:
P.2. Figure 1: Why focus on projected global mean sea level rise in the five countries with the largest share of global greenhouse gas emissions, rather than, for example, countries most at risk in terms of human populations or economic impacts?
P.3. line 40-45: consider clarifying a ‘qualitative state’. The intention here is to define social tipping as change to a qualitatively different system state, This could be made clearer.
P.3. Line 46 change to ‘enabling conditions’ typo inverted first quotation mark.
P.3. line 48 suggest “originating from either natural or social systems”.
P.3. line 48-49: I am unclear about the distinction between natural and social system forces on the one hand, and “those emerging from individual-level changes”. Are individuals not part of social systems?
P.3. Line 50-53: The authors claim that “Within social-ecological systems, experienced climate impacts (Demski et al., 2017; Konisky et al., 2016), e.g. floods and heat waves (Ricke and Caldeira, 2014), have the potential to shift attitudes and behaviors toward climate change and instigate social tipping processes (Müller et al., 2021).” There is evidence to support this claim. However, it is by no means certain that people affected by climate-related extreme events will be motivated to seek bolder climate action. Depending on a range of variables – including the kind and degree of affect/trauma, in-group influences, and individual differences – some people may become demotivated or use denial as a defence mechanism (Marshall, 2014). There is also evidence that increased societal stress and insecurity resulting from climate-related extreme weather events might also reduce tolerance and cooperation and make pro-environmental policies even less likely (Friedman, 2005). Forces of climate delay are also well resourced and creative in exploiting fear and loss. The authors might therefore consider the possibility that climate impacts might have the opposite effect of triggering and accelerating ‘damaging cascades’ (Lenton et al., 2022, p.8) of social unrest, populism, conflict and ‘barbarization (Raskin, 2016, p. 26), leading to a ‘fractured world’ (Laybourn-Langton, 2022, p. 10).
P.4. line 89. This section introduces the concept of complex contagion. But the ‘social activation model’ is not introduced or explained until the authors state that ‘the complex contagion social activation model of social tipping’ is to be applied in the model design. What is ‘the social activation model’? Is it a term coined by the authors? The Norm Activation Model (Schwartz, 1977), the Value-Belief-Norm Theory (Stern, 2000), and the Theory of Planned Behaviour (Ajzen, 1991) all offer explanations for how values, norms, beliefs, and intentions can motivate behavioural change Do the authors consider any of these models to be relevant in this case?
P.4. Line 93. The authors state that individual preference factors and network structures can ‘trigger rapid shifts in social norms and behaviors’. p. 5. Line 99: similarly, the authors identify ‘ rapid adoption of environmental behaviors’ . Do the authors have any evidence that changes in attitudes of the populations in the target countries are likely to experience an emergent threshold function (F(r(t))) as opposed to more incremental change dynamics?
P.5. Line 106 – spelling ‘heterogeneities’
In the authors methodological approach, they adopt a ‘low-dimensional’ approach to deal with the ‘uncertainties and heterogeneities’ of real-world populations, using a tripartite categorisation based on resource mobilization theory that is 40-50 years old. There is a lot more to climate action beyond resource mobilization. Why was this categorisation considered the most appropriate? And how does it affect the generalisability of the results?
p.10. line 230: the authors estimate their ‘potentially active population share’ for each of the target countries according to national survey data of climate change concern. However, a self-reported level of climate change concern does not necessarily denote any connection to the subjects’ awareness of elevation above sea level or to the level of risk the subjects face under different SLR projections.
Citation: https://doi.org/10.5194/egusphere-2023-1622-RC2 -
AC2: 'Reply to Reviewer 2', Keith Smith, 20 Mar 2024
We thank the reviewer for the overall positive review of our paper. We have responded to each of these comments and suggestions below, organised by topic.
[Why focus on 5 exemplary countries in Figure 1?]
The reason for focussing on the five countries with the largest shares of GHG emission in the first part of the manuscript’s results section is twofold: First, we acknowledge that those countries are potentially also among the ones where effective climate policy might have the most substantial impact on global carbon emissions. At the same time, it is reasonable to assume that these countries do in fact possess the political capacities to lead others towards joining into collaborative efforts to mitigate climate change. Second, quantifying which countries are most at risk with respect to human population or socio-economic impacts is a highly multidimensional problem that is still subject of current research. Had we attempted to put our focus according to any of these dimensions our choice would have likely become very subjective. Ultimately, we note that previous works on the interrelation between future climate variability and policy have opted for similarly focusing on the five top emitters (Ricke & Caldeira, 2014). For the revision we propose to add a statement that justifies our choice of presented countries using the arguments above to the beginning of Sec. 4.1 in the manuscript.
[Spelling and technical issues]
We agree with the reviewer that the notion of a qualitative state in social systems deserves further clarification.We therefore propose to add examples from technology adoption, political systems, and environmental behaviour to the respective section in the introduction of the paper. We thank the reviewer for highlighting a few typographic errors, which we will fix during the revision.
We will change the sentence on line 48 to ‘The ultimate trigger of social tipping processes can come from forces originating from either natural or social systems [...]’.
[Individual and societal level changes]
Yes, we agree that individuals are part of social systems. For this analysis, we focus on a theory of change that originates (is triggered) from individual (micro-) level dynamics, as opposed to from systemic level changes (e.g. policy change). That is, we explore how these individual changes can enact system regime changes, which themselves reshape the individual actions nested within these regimes. This approach is situated directly within the “social tipping” theory of change elaborated within the paper, but should naturally be interpreted as complimentary, and part of, broader mechanisms towards how such rapidly shifts could occur. The interplay between micro-macro-micro changes are a foundation of social analyses looking at the co-evolution of structure and agency (e.g. “Coleman’s Boat”, Coleman 1994). In a revised version of the manuscript, we would make the connection between micro actions and resulting macro level changes more clear and linking to relevant literature on such theories of change.
[Uncertainty of changes resulting from climate impacts]
This is a very good point that we agree with. As we note in our response to Reviewer #1 above, there is much hetereogeniety in the effect of climate impacts on climate change attitudes and behaviour, and as this comment correctly notes, not all of these changes may be normatively ‘positive’. Furthermore, there is recent literature exploring the risks of societal tipping dynamics (e.g. Milkoreit 2023), which emphasizes the inherent danger of rapid social change, which is a very chaotic event and the outcomes are far from determined. Indeed, contemporary examples used to identify social tipping (e.g. Arab Spring) have not always resulted in long-term, “positive”, change within these societies. We suggest revising the discussion and theoretical background to emphasize the uncertainty not only in the effects of extreme events, but the direction of the resulting changes.
[Model terminology and relation to other theories of environmental social change]
This is a good point that is also raised in regard to the relation of the theory of change in this model to other forms of change that have been applied to environmental behaviours. The current version of the manuscript does not clearly elaborate the connections between social tipping and established theories of change - as it is more focused on applying these concepts in the modelling exercise as opposed to theoretically explaining them. But here, we would highlight the direct connections between classical threshold models of behavioural change (e.g. Granovetter 1978, Schelling 1972) alongside those focused on developing a micro-based foundation for social change (as noted above, and in line with the VBN theory noted here). This is not as directly connected to TBP based mechanisms, or other theoretical explanations of the attitude-behaviours gap (e.g. Kollmuss and Agyemann, 2002; Diekmann and Preisendörfer, 2003), as these are rather understood by the authorship team as potential barriers and limitations to the theory of change explored by this model. We suggest revising the manuscript to make the connections between social tipping literature and other proposed theories of environmental change (individual and structural) more clear.
We also suggest changing the name of the model to be “network-based threshold model for social tipping” to further clarify these connections to previous literature (Wiedermann et al., 2020) and reduce confusion over this modelling approach.
Lastly, as noted in our response to Reviewer 1 [Definition of active population], we would also suggest changing the title to reflect the new terminology of the model, as well as the proposed change in terminology for climate action: “A global threshold model of enabling conditions for social tipping in pro-climate change behaviours – the role of sea-level rise anticipation and climate change concern”
[Connections to resource mobilization theory]
We agree with the reviewer that resource mobilization theory has already been around for half a century. In our original paper (Wiedermann et al., 2020) we motivate our choice of the model by acknowledging that spreading of opinions and behaviours can often be explained through complex contagion (Watts, 2002; Centola et al., 2015 & 2018). At the same time, recent studies have investigated the role of ‘instigators’ (Singh et al., 2013) or ‘immune’ individuals (Karsai et al., 2016) in addition to the contingent population, as such groups are commonly discussed in the literature and seem apparent in real-world situations. Notably, such distinctions, even though they were only studied quantitatively in the recent decade, align very well with the attribution of groups in resource mobilisation theory (RMT). In this sense, we only draw the terminology from RMT without inferring any causal mechanism related to that theory. We think that using a tripartite categorisation is appropriate as it approximates relevant actors in any social movement in a meaningful way without the need for specifying more groups than necessary: those that will act, those that are willing to act and those that will never act.
We propose to rework our manuscript so that it becomes clear that all we draw from RMT is the terminology of groups while the rest follows from literature that is much more recent.
[Connection between climate change concern and SLR perceptions]
We completely agree with the reviewer that concern can not necessarily be linked to awareness or proper risk assessment. However, it seems that there is a misunderstanding at play that we will attempt to clarify in the revision of the manuscript.
The core idea of differentiating between potentially active and certainly active individuals is that the former don’t necessarily need to be directly affected by sea-level rise. Rather, it is the certainly active population, i.e., those that are or will be affected, that we assume to trigger a movement so that the potentially active individuals will join in, regardless their specific locale in a country. In other words, in order to instigate a movement, an individual needs to be both affected and concerned. Only then will those that are concerned, but not affected, join into a certain climate action based on social contagion and peer pressure.
We also agree that concern is not even distributed within countries. Indeed, there may even be more intra-country variance than across countries. But, there is the potential for spillovers within a country, for example, a person living in Berlin could be very concerned about SLR (but not directly affected). Yet, the person is connected to an individual in Hamburg who is concerned and affected, which could trigger both towards action. We suggest including this as a limitation and suggestion for future research in the revised manuscript.
Further, this comment notes that we use climate change concern, and not SLR risk perceptions, to operationalise the potentially active population. This is rather a product of the data limitations, where questions about the perceived risks associated with climate change are often asked more broadly, and less frequently in respect to specific impacts (such as SLR). While we agree that we do not directly capture SLR perceptions in the data used for this study, climate change concern can function as an appropriate “best available data” proxy. For example, recent research conducted by the authorship team suggests a close relationship between anticipation of SLR and climate change concern, where we find a strong, positive correlation r=0.76 in a recent sample of US adults (see Figure A1 in Smith et al., 2022). Accordingly, we would add this information as part of the revised limitations mentioned above.
Citation: https://doi.org/10.5194/egusphere-2023-1622-AC2 -
AC5: 'References for all responses', Keith Smith, 20 Mar 2024
As a final note, please do find our references for these responses to all reviewers here:
Akerlof, K., Covi, M., & Rohring, E. (2017). Communicating sea level rise. In Oxford Research Encyclopedia of Climate Science.
Akerlof, K., Merrill, J., Yusuf, J. E., Covi, M., & Rohring, E. (2019). Key beliefs and attitudes
for sea-level rise policy. Coastal Management, 47(4), 406-428.
Bolsen, T., Kingsland, J., & Palm, R. (2018). The impact of frames highlighting coastal flooding in the USA on climate change beliefs. Climatic Change, 147, 359-368.
Coleman, J. S. (1994). Foundations of social theory. Harvard university press.
Covi, M. P., & Kain, D. J. (2016). Sea-level rise risk communication: Public understanding, risk perception, and attitudes about information. Environmental Communication, 10(5), 612-633.
Diekmann, A., & Preisendörfer, P. (2003). Green and greenback: The behavioral effects of environmental attitudes in low-cost and high-cost situations. Rationality and Society, 15(4), 441-472.
Kollmuss, A., & Agyeman, J. (2002). Mind the gap: why do people act environmentally and what are the barriers to pro-environmental behavior?. Environmental education research, 8(3), 239-260.
Marzeion, B., & Levermann, A. (2014). Loss of cultural world heritage and currently inhabited places to sea-level rise. Environmental Research Letters, 9(3), 034001.
Milkoreit, M. (2023). Social tipping points everywhere?—Patterns and risks of overuse. Wiley Interdisciplinary Reviews: Climate Change, 14(2), e813.
Müller, P. M., Heitzig, J., Kurths, J., Lüdge, K., & Wiedermann, M. (2021). Anticipation-induced social tipping: can the environment be stabilised by social dynamics?. The European Physical Journal Special Topics, 230(16-17), 3189-3199.
Priestley, R. K., Heine, Z., & Milfont, T. L. (2021). Public understanding of climate change-related sea-level rise. PLoS One, 16(7), e0254348.
Quoß, F., & Rudolph, L. (2022). Operationalisation matters: Weather extremes as noisy natural experiment show no influence on political attitudes. https://osf.io/preprints/osf/m5rz8
Smith, E. K., Eder, C., Donges, J. F., Heitzig, J., Katsanidou, A., Wiedermann, M., & Winkelmann, R. (2022). Domino Effects in the Earth System-The role of wanted social tipping points. OSF Preprints.
Smith, S. R. (2023). Enabling a political tipping point for rapid decarbonisation in the United Kingdom. EGUsphere, 2023, 1-21.
Tàbara, J. D., Frantzeskaki, N., Hölscher, K., Pedde, S., Kok, K., Lamperti, F., ... & Berry, P. (2018). Positive tipping points in a rapidly warming world. Current Opinion in Environmental Sustainability, 31, 120-129.
Tàbara, J. D. (2023). Regenerative sustainability. A relational model of possibilities for the emergence of positive tipping points. Environmental Sociology, 9(4), 366-385.
Thomas, M., Pidgeon, N., Whitmarsh, L., & Ballinger, R. (2015). Mental models of sea-level change: A mixed methods analysis on the Severn Estuary, UK. Global Environmental Change, 33, 71-82.
Wiedermann, M., Smith, E. K., Heitzig, J., & Donges, J. F. (2020). A network-based microfoundation of Granovetter’s threshold model for social tipping. Scientific reports, 10(1), 11202.
Citation: https://doi.org/10.5194/egusphere-2023-1622-AC5 -
AC6: 'Notes on organization of responses to reviewer comments', Keith Smith, 20 Mar 2024
First of all, we thank the reviewers for this constructive and well considered review of our manuscript. We will address these issues raised by each point below. In order to best organise our responses, we have structured these by Reviewer (underlined) and thematic concept being addresses [in brackets]. If similar issues are raised across reviewers, we address these more directly once, and cross-reference of our responses where appropriate for clarity.
Again, thank you all very much for these constructive and helpful feedback.
Citation: https://doi.org/10.5194/egusphere-2023-1622-AC6
-
AC2: 'Reply to Reviewer 2', Keith Smith, 20 Mar 2024
-
RC3: 'Comment on egusphere-2023-1622', Anonymous Referee #3, 24 Nov 2023
I really enjoyed reading the paper, it makes really a great and very much needed contribution to understanding mechanisms of positive social tipping. I also see that two reviews have already been published, so I'm not sure my review is even needed any longer and I will keep it brief.
In substantial terms my only main objection would be that the focus is on SLR, but climate change has a whole range of impacts (draughts, flooding, etc.) that will also affect populations that are land-locked or live more land inwards, this shortcoming becomes particularly apparent in the Tipping Class II. So maybe the authors could emphasise that this type of analyses should be expanded in future considering a range of climate changer impacts.
I also thought that the results on modular networks are quite interesting and worth to elaborate on. I think the suggestion that in the case of anticipation of SRL such modularised networks are less likely can be challenged. For instance, we know that there are inequalities (e.g. racial, class) at play when it comes to who is most likely to be affected by climate change impacts (incl. SLR) and in these circumstances a modularized network may actually be quite realistic and this result would then show how inequality could prevent tipping potentially?
Apart from that, I would urge the authors to go through the manuscript and check terminology and notation for consistency and clarity. For instance the reader may get confused between the notion of contingently active and potential active, as they signify very different population groups in the modelling, but this distinction is not clear enough in the chosen terminology.
I would also urge the authors avoiding using notations multiple times for different parameters etc. as this is extremely confusing. For instance s is used in equation 4 to represent a given survey in later equations it is used for simulation it seems and then in equation 9 for set of all ensemble members? u is used for upper branch but also in u_tot for tipping potential.
The explanation for in Figure 5 caption needs to come earlier, namely on p.11 (line 255).
In 3.5 state that tot stands for total and explain u_bif. The explanations are given only properly in Figure 5 caption.
Figure 5. You used United States in Figure 4, here you write United States of America, make sure this consistent.
Figure 6. I would suggest that the colours should be equivalent across the three images, otherwise it becomes very misleading as red may mean very different things in Figure A, B or C etc. Also add to B that this is total tipping potential.
Correct on p. 11 line 246, it should be Tab A2 and Tab A3.
Overall however, an excellent paper!
Citation: https://doi.org/10.5194/egusphere-2023-1622-RC3 -
AC3: 'Reply to Reviewer 3', Keith Smith, 20 Mar 2024
Thank you as well for these constructive comments, which we respond to by each topic raised below.
[SLR as chosen impact and distributional issues]
Thank you very much for the supportive comments. We very much agree with the potential to expand this analysis to other climate impacts. This is something we will further emphasise in the revised manuscript. In our response to the other expert reviews, we also note how we will further elaborate the rationale for why we focus on SLR.
We also agree with the need to focus on inequalities associated with these impacts. While the model does not allow for the specificity to talk about specific societal subgroups within a given country (e.g. race, class) - we can make a clear link between the likelihood of people to be exposed to SLR (and related impacts) and the comparative diminished capacity/agency of these people to respond. This would speak directly to the concern raised, that the people most likely to ‘tip’ are less likely to have the capacity to act. This is something that we view as important to highlight, and will expand upon in the revised discussion.
[Terminology and Specific Technical Changes]
We thank you for these more specific comments as well. We will address each of these accordingly within the revised manuscript.
Citation: https://doi.org/10.5194/egusphere-2023-1622-AC3 -
AC5: 'References for all responses', Keith Smith, 20 Mar 2024
As a final note, please do find our references for these responses to all reviewers here:
Akerlof, K., Covi, M., & Rohring, E. (2017). Communicating sea level rise. In Oxford Research Encyclopedia of Climate Science.
Akerlof, K., Merrill, J., Yusuf, J. E., Covi, M., & Rohring, E. (2019). Key beliefs and attitudes
for sea-level rise policy. Coastal Management, 47(4), 406-428.
Bolsen, T., Kingsland, J., & Palm, R. (2018). The impact of frames highlighting coastal flooding in the USA on climate change beliefs. Climatic Change, 147, 359-368.
Coleman, J. S. (1994). Foundations of social theory. Harvard university press.
Covi, M. P., & Kain, D. J. (2016). Sea-level rise risk communication: Public understanding, risk perception, and attitudes about information. Environmental Communication, 10(5), 612-633.
Diekmann, A., & Preisendörfer, P. (2003). Green and greenback: The behavioral effects of environmental attitudes in low-cost and high-cost situations. Rationality and Society, 15(4), 441-472.
Kollmuss, A., & Agyeman, J. (2002). Mind the gap: why do people act environmentally and what are the barriers to pro-environmental behavior?. Environmental education research, 8(3), 239-260.
Marzeion, B., & Levermann, A. (2014). Loss of cultural world heritage and currently inhabited places to sea-level rise. Environmental Research Letters, 9(3), 034001.
Milkoreit, M. (2023). Social tipping points everywhere?—Patterns and risks of overuse. Wiley Interdisciplinary Reviews: Climate Change, 14(2), e813.
Müller, P. M., Heitzig, J., Kurths, J., Lüdge, K., & Wiedermann, M. (2021). Anticipation-induced social tipping: can the environment be stabilised by social dynamics?. The European Physical Journal Special Topics, 230(16-17), 3189-3199.
Priestley, R. K., Heine, Z., & Milfont, T. L. (2021). Public understanding of climate change-related sea-level rise. PLoS One, 16(7), e0254348.
Quoß, F., & Rudolph, L. (2022). Operationalisation matters: Weather extremes as noisy natural experiment show no influence on political attitudes. https://osf.io/preprints/osf/m5rz8
Smith, E. K., Eder, C., Donges, J. F., Heitzig, J., Katsanidou, A., Wiedermann, M., & Winkelmann, R. (2022). Domino Effects in the Earth System-The role of wanted social tipping points. OSF Preprints.
Smith, S. R. (2023). Enabling a political tipping point for rapid decarbonisation in the United Kingdom. EGUsphere, 2023, 1-21.
Tàbara, J. D., Frantzeskaki, N., Hölscher, K., Pedde, S., Kok, K., Lamperti, F., ... & Berry, P. (2018). Positive tipping points in a rapidly warming world. Current Opinion in Environmental Sustainability, 31, 120-129.
Tàbara, J. D. (2023). Regenerative sustainability. A relational model of possibilities for the emergence of positive tipping points. Environmental Sociology, 9(4), 366-385.
Thomas, M., Pidgeon, N., Whitmarsh, L., & Ballinger, R. (2015). Mental models of sea-level change: A mixed methods analysis on the Severn Estuary, UK. Global Environmental Change, 33, 71-82.
Wiedermann, M., Smith, E. K., Heitzig, J., & Donges, J. F. (2020). A network-based microfoundation of Granovetter’s threshold model for social tipping. Scientific reports, 10(1), 11202.
Citation: https://doi.org/10.5194/egusphere-2023-1622-AC5 -
AC6: 'Notes on organization of responses to reviewer comments', Keith Smith, 20 Mar 2024
First of all, we thank the reviewers for this constructive and well considered review of our manuscript. We will address these issues raised by each point below. In order to best organise our responses, we have structured these by Reviewer (underlined) and thematic concept being addresses [in brackets]. If similar issues are raised across reviewers, we address these more directly once, and cross-reference of our responses where appropriate for clarity.
Again, thank you all very much for these constructive and helpful feedback.
Citation: https://doi.org/10.5194/egusphere-2023-1622-AC6
-
AC3: 'Reply to Reviewer 3', Keith Smith, 20 Mar 2024
-
RC4: 'Comment on egusphere-2023-1622', Anonymous Referee #4, 05 Feb 2024
This is a nice paper in that it examines a potential mechanism by which positive social tipping might come about – specifically, using a social activation model to explain increased concern as a function of (anticipation of) sea level rise. I had a few questions / suggestions that might strengthen the paper or bring it in line with some existing literature.
First, I wondered why the authors focused on SLR rather than extreme weather events that are happening now/on a much shorter timescale, such as hurricanes, wildfires, or droughts? Much of the literature on concern as a result of experience with climate change is focused on extreme weather events, rather than SLR, which is a much slower and distant process (with long anticipation horizons) – due to this prolonged time horizon, people may also expect that they/societies will be able to adapt. A focus on extreme weather impact would link the paper to this literature and provide a clearer basis for the increased concern the authors suppose.
Second, I was not familiar with the social activation model. However, the pathway from concern to active population or action was not sufficiently motivated. Concern is, on its own, unlikely to lead to action – many people globally are currently concerned about climate change and there is still insufficient action, and concern need not be a precursor for action (i.e. action may be a beneficial proposition in its own right). Much of the literature on experience of extreme weather and attitudes and behaviors finds tenuous relationship with behavior, and a much more robust relationship to concern. Even in areas with regular flooding and extreme weather events (e.g. Florida, Texas Gulf Coast), there is limited concern about climate change. The authors should be clear about this limitation – focus on concern rather than actions/behavior - up front. They could frame the paper in terms of a tipping point in concern rather than action (as it is currently unclear what the tipped state/actions are), which would require a more nuanced explanation of how one begets the other (which will vary across context, depend on government action, etc.). In the discussion, they could then speculate that growing concern could lead to subsequent action through pressure on politicians etc. Currently there is some discussion in the introduction, but it then seems lost throughout. Also, action is a broad term – it is unclear which actions the authors are referring to… is this mitigative actions? Adaptation? Finally, responses are likely to vary substantially across countries due to differences in income/adaptive capacity/culture etc. These may also very well be correlated with SLR exposure. It would be great if the authors could account for some of this heterogeneity.
The authors argue that for the case of SLR anticipation, real-life social networks are unlikely to have modularized network structures, but I am not sure why this is the case. For example, there are very different demographics / types of jobs etc. in rural vs coastal areas. There is literature showing that people sort in where they live – often choosing to leave nearby close others (and migrating to areas where they know people). In the US, many exposed areas also have climate skeptics / deniers / and Republicans, who are less likely to endorse climate action. Relatedly, the authors do not discuss the literature on motivated reasoning but there is some literature showing that people often interpret extreme weather events in ways that fit their prior beliefs – these could be the never active types, but they are likely also some of the potentially active types. In other words, the same amount of information might result in heterogeneous responses depending on prior beliefs even among potentially active types.
It was not clear to me why the authors exclude low elevation regions. Do they expect concern to propagate in a different way in these areas?
The discussion of sensitive interventions, with mentions of policy regime changes, social movements, policy entrepreneurs etc., comes a bit out of the blue here and seems non-specific / lacking nuance relative to the rest of the piece. These sections seem a little tacked on, and the mechanisms aren’t fleshed out so they seem a bit unsatisfying.
Citation: https://doi.org/10.5194/egusphere-2023-1622-RC4 -
AC4: 'Reply to Reviewer 4', Keith Smith, 20 Mar 2024
Lastly, we greatly appreciate the feedback provided as well. Similar to the other comments, we have responded by topic below, as well with linkages to other responses to similar comments we have addressed above.
[SLR as Chosen Climate Impact]
Thank you very much for this comment. This is something that is shared commonly across many of the reviewers. For the sake of parsimony, we will not repeat our responses here, but would rather refer to our replies to Reviewer #1 [Why does the paper focus on SLR and not on other impacts? Comment #1] and [Why does the paper focus on SLR and not on other impacts? Comment #2] and to our response to Reviewer #2 [Uncertainty of changes resulting from climate impacts].
[Heterogeneity in the effect of concern]
Thank you for raising this point, we very much agree with this. We also note some of these issues about the variance in the effect of concern in shaping climate actions in our response to Reviewer #1 [Salience of SLR and how SLR perceptions shape attitudes and behaviors] and Reviewer #2 [Connection between climate change concern and SLR perceptions] (amongst other comments above). We very much appreciate the further suggestion to expand the discussion on what this means in terms of political relevance. If a system has higher levels of concern, we find that it has the necessary conditions for change, and in this case, other efforts (such as pressuring politicians) could be successful to push towards a tipped state. But of course, what action can be taken to “trigger” the social tipping depends on the local context. We suggest expanding the discussion in the revised manuscript to make these recommendations and relevance more clear.
[Defining Action]
Thank you for this comment as well, this was similarly shared by Reviewer #1 [Definition of active population]. We will refine this definition in the revised manuscript in line with our response to Reviewer #1, but in a particular response to this comment, we would consider adaptation to SLR to likely not be included as action, as in this context, this likely means that the impact of SLR has already occurred, and is likely unchangeable during the normal lifetime of humans. Rather, we are interested in how people can be motivated to act to minimise the risks of being exposed to SLR - so focusing on mitigating actions.
[Heterogeneity across subgroups within population]
We also agree with some of the concerns raised here about how the mechanism of never active->potentially active->certainly active may vary for certain subgroups within a population. Given the model resolution, such intra-country dynamics are not included in these analyses. Certainly, these could work both ways - where some countries will have sub-groups that are more likely to act and some that are not. We agree that this is a limitation that would be discussed in a future revision of the paper, as well as a call for further research into within country dynamics, which could build upon this assessment.
[Exclusion of lower elevation regions]
We had excluded lower lying regions (e.g. Azerbijian, Kazakhstan and the Netherlands) as well as those that are situated inland (e.g. Mongolia, Austria) as these present more ‘extreme’ case that are likely far outside of data distribution of the remaining sample. Particularly in the case of lower-lying countries, we consider these to be more specific cases, that are not likely as well explained by the model. This is an approach that has also been adopted by similar modelling exercises (e.g. Marzeion and Levermann, 2014). We would suggest revising the manuscript to make this justification more clear, and if the reviewer believes appropriate, present additional analyses including the data from lower lying regions as a robustness check.
[Connection to theories of change]
Thank you for the final comment on the discussion of sensitive interventions and other theories of change. This is similar to a comment raised by Reviewer #1 [Model terminology and relation to other theories of environmental social change]. Accordingly, we suggest revising this section in the discussion, to be more clear how social tipping relates to other theories of change, and how the proposed change would occur (bottom-up). We believe this would make these statements more clear to the reader.
Citation: https://doi.org/10.5194/egusphere-2023-1622-AC4 -
AC5: 'References for all responses', Keith Smith, 20 Mar 2024
As a final note, please do find our references for these responses to all reviewers here:
Akerlof, K., Covi, M., & Rohring, E. (2017). Communicating sea level rise. In Oxford Research Encyclopedia of Climate Science.
Akerlof, K., Merrill, J., Yusuf, J. E., Covi, M., & Rohring, E. (2019). Key beliefs and attitudes
for sea-level rise policy. Coastal Management, 47(4), 406-428.
Bolsen, T., Kingsland, J., & Palm, R. (2018). The impact of frames highlighting coastal flooding in the USA on climate change beliefs. Climatic Change, 147, 359-368.
Coleman, J. S. (1994). Foundations of social theory. Harvard university press.
Covi, M. P., & Kain, D. J. (2016). Sea-level rise risk communication: Public understanding, risk perception, and attitudes about information. Environmental Communication, 10(5), 612-633.
Diekmann, A., & Preisendörfer, P. (2003). Green and greenback: The behavioral effects of environmental attitudes in low-cost and high-cost situations. Rationality and Society, 15(4), 441-472.
Kollmuss, A., & Agyeman, J. (2002). Mind the gap: why do people act environmentally and what are the barriers to pro-environmental behavior?. Environmental education research, 8(3), 239-260.
Marzeion, B., & Levermann, A. (2014). Loss of cultural world heritage and currently inhabited places to sea-level rise. Environmental Research Letters, 9(3), 034001.
Milkoreit, M. (2023). Social tipping points everywhere?—Patterns and risks of overuse. Wiley Interdisciplinary Reviews: Climate Change, 14(2), e813.
Müller, P. M., Heitzig, J., Kurths, J., Lüdge, K., & Wiedermann, M. (2021). Anticipation-induced social tipping: can the environment be stabilised by social dynamics?. The European Physical Journal Special Topics, 230(16-17), 3189-3199.
Priestley, R. K., Heine, Z., & Milfont, T. L. (2021). Public understanding of climate change-related sea-level rise. PLoS One, 16(7), e0254348.
Quoß, F., & Rudolph, L. (2022). Operationalisation matters: Weather extremes as noisy natural experiment show no influence on political attitudes. https://osf.io/preprints/osf/m5rz8
Smith, E. K., Eder, C., Donges, J. F., Heitzig, J., Katsanidou, A., Wiedermann, M., & Winkelmann, R. (2022). Domino Effects in the Earth System-The role of wanted social tipping points. OSF Preprints.
Smith, S. R. (2023). Enabling a political tipping point for rapid decarbonisation in the United Kingdom. EGUsphere, 2023, 1-21.
Tàbara, J. D., Frantzeskaki, N., Hölscher, K., Pedde, S., Kok, K., Lamperti, F., ... & Berry, P. (2018). Positive tipping points in a rapidly warming world. Current Opinion in Environmental Sustainability, 31, 120-129.
Tàbara, J. D. (2023). Regenerative sustainability. A relational model of possibilities for the emergence of positive tipping points. Environmental Sociology, 9(4), 366-385.
Thomas, M., Pidgeon, N., Whitmarsh, L., & Ballinger, R. (2015). Mental models of sea-level change: A mixed methods analysis on the Severn Estuary, UK. Global Environmental Change, 33, 71-82.
Wiedermann, M., Smith, E. K., Heitzig, J., & Donges, J. F. (2020). A network-based microfoundation of Granovetter’s threshold model for social tipping. Scientific reports, 10(1), 11202.
Citation: https://doi.org/10.5194/egusphere-2023-1622-AC5 -
AC6: 'Notes on organization of responses to reviewer comments', Keith Smith, 20 Mar 2024
First of all, we thank the reviewers for this constructive and well considered review of our manuscript. We will address these issues raised by each point below. In order to best organise our responses, we have structured these by Reviewer (underlined) and thematic concept being addresses [in brackets]. If similar issues are raised across reviewers, we address these more directly once, and cross-reference of our responses where appropriate for clarity.
Again, thank you all very much for these constructive and helpful feedback.
Citation: https://doi.org/10.5194/egusphere-2023-1622-AC6
-
AC4: 'Reply to Reviewer 4', Keith Smith, 20 Mar 2024
Status: closed
-
RC1: 'Comment on egusphere-2023-1622', Anonymous Referee #1, 28 Aug 2023
General Comments
The manuscript is well-written and addresses the very important topic of better understanding enabling conditions for social tipping towards climate action. Quantitative approaches for socio-ecological systems are urgently needed to complement existing research for societal transformation towards climate resilience in its many facets. As this transformation is urgent and requires broad societal agreement, understanding the pathways for shift perception into an aware and ‘climate active’ state is timely and this approach provides a valuable contribution in this regard.
While the framing and approach are valuable, there are a number of points that would require additional work and analysis for the paper to achieve its full potential.
One of the key issues I see with the paper is that it focusses on one very important and also detrimental climate impact, but one that is very localised in that it applies to coastal regions only. While these are home to large cities and a large share of the global population, including in higher density settings, nonetheless it is for many regions not one of the most pressing impact and not as globally applicable (as the authors also point out in their manuscript). The anticipation horizon, as also discussed and assessed in the paper, is another key issue with SLR that, in my view, does not make it the ideal impact to focus on.
Another concern relates to the idea of ‘active population’ and how these active groups then lead to the transformative shift in climate action. I think more granularity and context is needed here. Concern does not necessarily translate into action and active populations don’t necessarily trigger the needed policy-shift.
Some of the more specific comments below provide further granularity to these to concerns, which hopefully help in revising the manuscript.
Specific comments
There are a number of globally relevant impacts that are already detrimental today and affect people globally. The most detrimental effects of SLR are very long-term and thus often operate on timescales beyond human comprehension (see Figure 1, for example), as the authors also point out in the para p2-3 l30-35. Those impacts of SLR felt today are often very localised to smaller coastal communities and even in highly affected areas today, awareness of cc let alone activism is not present. It is not sufficiently clear to me why SLR would be chosen as the relevant impact variable, when other impacts might be much more central in the current perception. As the authors point out on p3 l50-55, “Experienced climate impacts such as floods and heat waves have the potential to shift attitudes and behaviours towards climate change and instigate social tipping processes”. Why not focus on one of these widely experienced impacts for the model?
In the end, which impact actually triggers the enabling conditions for mitigation doesn’t matter as the result is the same across impacts. I would therefore disagree with the statement on p3 l73/74. In addition, the cited reference from 2012 seems both outdated in light of currently experienced impacts and the paper also does not seem to substantiate the claim made in the sentence.
It is not quite clear what is defined as ‘active population’ and how that relates to mitigation action. Of course, awareness of the population is an essential ingredient, but awareness is far from activism, let alone government follow-through, which would be what is needed to actually mitigate in line with the Paris Agreement. An EIB study found that for Germany, for example, 63% of the population in 2021 would favour stricter climate measures. According to my reading of “active” within the study, this would then be the majority of the German population. However, we do not see a tipping in mitigation action.
While I see the importance of the model developed in this study, I think it needs more nuancing in terms of what is can actually show and how this activation of tipping then leads to large-scale global action. Maybe this would be better placed within the specific environment of policy-makers and how tipping can be induced with them?
I would also like to get more context for the consideration of ‘certainly active’ population as defined on p.6 line 129. While I see the point of assuming these should be ‘certainly active’ in terms of climate awareness, this is not necessarily the case for various reasons (e.g. US in Florida; West Africa). There are certainly no large activist movements there, despite the obvious current impacts and future exposure.
Maybe it would be worth further specifying (and calibrating) the model around a specific type of climate engagement within a specific group to first understand the context under which these groups have been activated to then be able to extrapolate to that specific type of action. The extension by which the activated population would then be able to force policy-action would then need to be an additional (albeit essential) extension of the analysis.
As mentioned above, I think it is critical to further define what is meant by ‘active’. Large parts of the population are convinced that climate action is needed, but are not openly active about it. How would these be placed within the model? In this context, adding different degrees of ‘active’ – along the lines of ‘aware’ to ‘willing to engage in existing activism’ to ‘active’ might be useful to then understand how activism may trickle down. Simply having an ‘aware’ population (as seems to be the definition of active in this paper) does not lead to the action that would be need to address the climate challenge. The authors provide quite a nuanced discussion of this in Sec. 1.1, but I’m not sure I see this reflected in the final assessment. Linking back to my comment above, this could also be linked to the reflection around which climate impacts drive awareness and action and create the enabling conditions for change.
Finally, I think the discussion of enabling conditions (line 45 and following) might require the consideration of some additional literature that provides quite a lot of a additional granularity (see e.g. IPCC WG2 Ch17 on adaptation and risk or IPCC WG3 Ch17 on Transitions). Clearly for a modelling approach, some reduction of complexity is needed, but I feel that inclusion of some of this complexity, including interactions between different aspects of enabling conditions, would be needed. As per some of my comments below, I feel results to some extent appear obvious out of the model set-up and if this complexity is not accounted for, I’m not convinced a model is actually needed for the results we get.
Results
- 12 lines 287-289 isn’t this conclusion rather obvious from the model set-up?
- 14 lines 320/325 These are super interesting results and also point towards what I mentioned in the General Comments above: there are different types of impacts that would most likely be a good predictor of climate action. In the Pacific with above average SLR rates, SLR may well be the perceived most pressing impact (though tropical storms may well contribute to the awareness as well). It may be interesting to choose regionally specific impact drivers that better reflect regionally specific risks , as this would likely provide a much closer to reality situation.
Discussion
The discussion raises a number of important points and also highlights the importance of better understanding social tipping processes to target interventions where they are most likely to yield results. Exactly due to the importance of such work, I would strongly encourage a further sharpening of the analysis to be more directly relevant for understanding the complexities.
In summary, as outlined above, on the one hand, I think the approach would strongly benefit from considering further key impacts that drive awareness globally. Unfortunately, the recent years have given plenty of examples of what these may be (wildfires, droughts, heatwaves, flooding…). On the other hand, more nuancing of what ‘active’ means and how this translates into the needed policy action would be important to include.
Citation: https://doi.org/10.5194/egusphere-2023-1622-RC1 -
AC1: 'Reply to Reviewer 1', Keith Smith, 20 Mar 2024
Thank you very much for these comments and constructive feedback on the manuscript. We have grouped each of our responses thematically [listed in brackets] and have also linked to where we have responded to similar comments by other expert reviewers as well. Again, thank you again very much for the efforts and support of this manuscript.
[Why does the paper focus on SLR and not on other impacts? Comment #1]
One of the issues raised is that the analysis of this paper focuses on sea-level rise, as SLR is particularly localised. This concern is valid, and well-taken by the authorship team, but this is also true of most other projected future climate impacts. For example, floods, droughts, increased intensity and frequency of hurricanes likely share similar spatial impacts. This comment also raised the issue of the temporal distance between mitigating actions and the manifestation of SLR. This is particularly the motivating reason why we chose to focus this empirical modelling work on SLR, as this is an impact in which there is limited observed data. Rather, the primary question of this research is how a problem like anticipation of future SLR can (or cannot) lead towards increased climate mitigating behaviours and policies. There are many other papers which focus on the impact of extreme weather events (often using observed data after exposure to the event) - but for the future impacts of SLR on climate actions, this is something that can likely only be assessed via a modelling exercise such as this one (particularly on the global scale). Further to this point, we believe that SLR is a particularly unique potential climate impact not only because of the temporal disconnect between mitigating actions now are need to minimise impacts that will manifest in the future, but also importantly, SLR is a very unique potential impact of climate change, as it is particularly persistent. The impact of floods (for example), while potentially causing similar amounts of damage to localised communities and land, are comparatively more temporary. Indeed, we see that in many areas, communities rebuild on the same flood lands within comparatively short periods of time (years). While for SLR, the impacts persist over potentially thousands of years - which for current human societies is effectively permanent. In this case, the impact of SLR is a very risky form of climate change induced event, and one that is comparatively less understood within the social science and empirical literature. Accordingly, we suggest that the revised manuscript would further develop the argument for why SLR is a unique and important climate change induced impact, as well as further developing the existing literature on other impacts to note the differences and similarities with SLR.
[Theory of change and model specification - how do people translate into active population]
We thank the reviewer for bringing up these questions about the ‘active’ population as well. We agree that concern does not necessarily translate into action, this is similar to the problem of an ‘attitudes-behaviour’ gap. Our model intends to incorporate this uncertainty, where concern is rather viewed as a necessary - but not sufficient - condition for action. We parameterise the ‘potentially active’ population using the level of concern for a given country - which simply means that the person has the potential to become activated, but would only do so if a majority of their neighbour nodes become activated. This is clearly reducing a great amount of complexity in human behaviour - but given the model resolution, we explicitly do not intend to deterministically explain how people will become active. Still, we note that such an approach is in line with a comparatively large body of literature on models of social activation that effectively group actors into similar categories like the ones we employ here. Given that our model explores the complexities of the relationship between individual attitudes, network effects, SLR exposure and impact time, we keep the level of specificity for human behaviour as relatively low, as otherwise, the model components would be very difficult to identify and parameters difficult, if not impossible, to estimate. Thus, human behaviours would likely be at a different depth of specification and out of line with the other model components.
[Why does the paper focus on SLR and not on other impacts? Comment #2]
We thank you for the further comment regarding the choice of SLR as the chosen climate impact we assess in this model. Similar to our response above, we note that SLR is a unique, and comparatively understudied future climate impact. There is a comparatively large literature on the role of extreme weather events affecting attitudes and behaviours - yet even within this literature, the effects vary substantially by operationalization of the independent (event) and dependent (attitude and behaviour). For a recent finding in this regard, Figure 6 and 7s of this pre-print from Quoss and Rudolph (2023) nicely demonstrate substantial noise in the impact of extreme temperature, precipitation and other events shaping attitudes in Switzerland.
That is, we also recognise the difficulty in determining what will actually “cause” social tipping, and in a revised manuscript, would further note the limitations of any extreme event in initiating tipping dynamics - as there is limited empirical evidence that this has yet to occur.
At the same time, we further believe that SLR is an interesting case study, with the potential to motivate action - even though the impacts are indeed likely (hopefully) quite far into the future. For example, a related paper by the authorship group (under review) uses survey data from the United States (a locale with comparatively lower proportion of citizens likely to be affected directly by SLR - Class III), finding that concern about SLR is one of the top factors driving climate change policy support and behavioural willingness (see Smith et al., 2022).
Accordingly, this paper focuses on understanding how SLR would link current actions with impacts that will not manifest until generations into the future via the mechanism of anticipation. Indeed, this may be an optimistic - but important question - as actions in response to experienced SLR would already be far too late to substantially mitigate these quasi-permanent changes to vast sections of inhabited land around the world.
Lastly, in this paper we do propose a linkage between SLR and other climate impacts, where anticipation of SLR can provide grounds for event-induced tipping, i.e., the anticipation of SLR can raise concern sufficiently high so that other, more immediate impacts, can kick the system into a new stable state.
[Salience of SLR and how SLR perceptions shape attitudes and behaviors]
This point is very well taken regarding the claim made about the salience of SLR and the reference citation. We have further reviewed the literature on the salience and impact of sea-level rise. Several studies have noted that sea level rise is not particularly well understood by the public (in terms of scientific accuracy, e.g. Thomas et al., 2015, Priestley et. al, 2021), is seen as occurring quite far into the future (Covi et al., 2016), and some studies have also found that SLR does not receive frequent media attention (e.g. Akerloff et al., 2017; 2019). Yet, communicating the risks of SLR can increase climate change attitudes - even amongst more ‘hard to reach’ audiences such as Republicans in the United States (Bolsen et al., 2018; Smith et al., 2022). Accordingly, we suggest revising the manuscript to further include these relevant literatures, and develop a similar, but re-framed argument:
While SLR may be, at times, be misunderstood and not a priority for some members of the public, it has the potential to be a strong driver of climate change attitudes and behavioral change. Yet, the impact of anticipation of SLR on climate change attitudes and behaviours may be limited if the distance to projected impacts extends far into the future. Accordingly, we model how the likelihood of attitudes and behavioral change relates to differing levels of concern, being spread throughout networked dynamics, at varying levels of impact severity and timeframes.
[Definition of active population]
We agree with this comment and those further raised below - a similar concern to those brought by Reviewer 2 and 4 - regarding the lack of definition for the construct ‘active population’. We intended to draw upon an (intentionally simplified) categorization of the population into three groups - certainly active, potentially active, and not active. For the active population in this regard, these would be people that are already engaging to mitigate future climate impacts in their local settings. Given the cross-national design, the ‘action’ is relative to the needs and capacity for a given setting. This could be political action (e.g. voting behaviour directly for climate change policies, protesting), individual mitigating behaviours (e.g. pro-climate behaviours), or even social actions (spreading of climate communication, shifting norms). Given the resolution and cross-national implementation of this model, we cannot define more specifically what ‘action’ means, as any given action for any given context and person will a diverse set of motivations and barriers (complexity would need to be explored within a different empirical approach). Furthermore, we also note that the capacity to action for a given individual is likely inverse to the expected impacts of SLR - a distributional inequity posed by climate change. We would propose further revising the manuscript to better elaborate on the definition of climate action, and these distributional impacts.
Furthermore, we also suggest changing the terminology to be more reflective of the social science literature. The most broad phrase that is commonly used is ‘pro-climate change behaviors’. Accordingly, would suggest changing the terminology from climate action to ‘pro-climate change behaviors’ to be more reflective of these literatures.
Also, given the concerns regarding the terminology used for the modelling approach raised by Reviewer 2 [Model terminology and relation to other theories of environmental social change], we suggest revising the title to be “A global threshold model of enabling conditions for social tipping in pro-climate change behaviours – the role of sea-level rise anticipation and climate change concern”
We would further suggest that future research should be engaged to elaborate the mechanisms by which action can develop. Such work would need to be well-tailored to the individual action and setting. There is much work in the regards (largely enacted by environmental social psychologists, public policy researchers and behavioural economists) focusing on specific settings and actions - but disproportionately focusing on Western, wealthier states. Our goal of this modelling exercise was to intentionally to be rather abstract - allowing for cross-national comparisons. But these findings should be seen as in compliment to this rich literature exploring the more specific and diverse mechanisms of environmental change (individually, socially and politically).
Lastly, in regard to the EIB study regarding the desire of 63% of Germans to prefer stricter climate actions, we would rather consider this to be akin to the potentially active population, those willing to act, rather than to the actually active population. Responding positively to a survey question more likely displays a willingness to change, while the actual observed action would be whether this translates into political action (voting). And, as this comment notes, we did not observe tipping like behaviour within the German system (at least not yet). For example, there was a historic increase in support for the Green Party in the 2021 German Federal Election - where climate change was one of the largest issues across all parties. Yet, in the end, the Green Party only received ~15% of the popular vote and seats in Parliament. This resulted in the Green Party being part of the governing coalition (for the first time in 15 years), and likely contributed to many federal actions to combat climate change, but did indeed fall short of rapid, systemic changes. We would read this as evidence of a system that is getting closer towards change, but one that has, as of yet, not tipped.
[Specification of model and policy relevance of findings - Comment #1]
As noted in the previous comment, we also agree with the importance of developing specific mechanisms for how tipping can be induced, and proscribing actions for policy-makers and stakeholders to incentivise these changes. This is important future work - empirical research that would rely upon observed individual data (potentially within experimental settings) to understand how such changes can occur for a given context and towards a specific change. For example, there is another pre-print that focuses on expert elicitations of tipping dynamics towards decarbonization in the UK political system (Smith 2023). Yet, we would also note that defining such a specific mechanism is not the purpose or design of our research question and modelling exercise. This would require a far greater level of specificity, using different methods, and would almost certainly not be generalizable across such a cross-national design. We would revise our future research statement in a revised manuscript to make this call for further research more clear and direct.
Yet, in terms of political relevance from our modelling approach, we emphasise the role of creating enabling conditions for change to occur (e.g. Tabara et al., 2018; 2023). We find, that on it’s own, concern may be insufficient to induce tipping dynamics, but rather, provides the necessary conditions for such to change to take place (either induced by other events, policy changes, action etc). That it, concern pushes the system closer towards the threshold, such that tipping can occur.
[Specification of model and policy relevance of findings - Comment #2]
We further thank you for the comments regarding the specificity and complexity of this model. As we have noted above, we very much agree with the need for further research (across many disciplines) that further explores this complexity. For example, within a modelling exercise, research on interactions using a similar modelling design could also be engaged. Members of the authorship team have also engaged in such work using a similar model (e.g. Müller et al., 2021). We will further elaborate on different types of future research that could expand upon these findings (from an experimental, observational and modelling approach) within the revised manuscript.
We would also contend that, like most models, some of the conclusions are more directly related to the modelling design (that is, the more people are exposed to direct impacts of SLR the more likely they are to act). Yet, we contend that others are not as “obvious”. For example, the differing role of concern in shaping the likelihood of action. This is illuminated within the three classes of tipping section of the results - where we find that yes, within class I, the results are rather expected, but in classes II and III, there are interesting interactions between SLR and concern that are exposed, which we believe are non-trivial and not necessarily anticipated. We would suggest revising the manuscript to further emphasise the results that go beyond the “top line” findings, that may be more related to the modelling design itself.
Citation: https://doi.org/10.5194/egusphere-2023-1622-AC1 -
AC5: 'References for all responses', Keith Smith, 20 Mar 2024
As a final note, please do find our references for these responses to all reviewers here:
Akerlof, K., Covi, M., & Rohring, E. (2017). Communicating sea level rise. In Oxford Research Encyclopedia of Climate Science.
Akerlof, K., Merrill, J., Yusuf, J. E., Covi, M., & Rohring, E. (2019). Key beliefs and attitudes
for sea-level rise policy. Coastal Management, 47(4), 406-428.
Bolsen, T., Kingsland, J., & Palm, R. (2018). The impact of frames highlighting coastal flooding in the USA on climate change beliefs. Climatic Change, 147, 359-368.
Coleman, J. S. (1994). Foundations of social theory. Harvard university press.
Covi, M. P., & Kain, D. J. (2016). Sea-level rise risk communication: Public understanding, risk perception, and attitudes about information. Environmental Communication, 10(5), 612-633.
Diekmann, A., & Preisendörfer, P. (2003). Green and greenback: The behavioral effects of environmental attitudes in low-cost and high-cost situations. Rationality and Society, 15(4), 441-472.
Kollmuss, A., & Agyeman, J. (2002). Mind the gap: why do people act environmentally and what are the barriers to pro-environmental behavior?. Environmental education research, 8(3), 239-260.
Marzeion, B., & Levermann, A. (2014). Loss of cultural world heritage and currently inhabited places to sea-level rise. Environmental Research Letters, 9(3), 034001.
Milkoreit, M. (2023). Social tipping points everywhere?—Patterns and risks of overuse. Wiley Interdisciplinary Reviews: Climate Change, 14(2), e813.
Müller, P. M., Heitzig, J., Kurths, J., Lüdge, K., & Wiedermann, M. (2021). Anticipation-induced social tipping: can the environment be stabilised by social dynamics?. The European Physical Journal Special Topics, 230(16-17), 3189-3199.
Priestley, R. K., Heine, Z., & Milfont, T. L. (2021). Public understanding of climate change-related sea-level rise. PLoS One, 16(7), e0254348.
Quoß, F., & Rudolph, L. (2022). Operationalisation matters: Weather extremes as noisy natural experiment show no influence on political attitudes. https://osf.io/preprints/osf/m5rz8
Smith, E. K., Eder, C., Donges, J. F., Heitzig, J., Katsanidou, A., Wiedermann, M., & Winkelmann, R. (2022). Domino Effects in the Earth System-The role of wanted social tipping points. OSF Preprints.
Smith, S. R. (2023). Enabling a political tipping point for rapid decarbonisation in the United Kingdom. EGUsphere, 2023, 1-21.
Tàbara, J. D., Frantzeskaki, N., Hölscher, K., Pedde, S., Kok, K., Lamperti, F., ... & Berry, P. (2018). Positive tipping points in a rapidly warming world. Current Opinion in Environmental Sustainability, 31, 120-129.
Tàbara, J. D. (2023). Regenerative sustainability. A relational model of possibilities for the emergence of positive tipping points. Environmental Sociology, 9(4), 366-385.
Thomas, M., Pidgeon, N., Whitmarsh, L., & Ballinger, R. (2015). Mental models of sea-level change: A mixed methods analysis on the Severn Estuary, UK. Global Environmental Change, 33, 71-82.
Wiedermann, M., Smith, E. K., Heitzig, J., & Donges, J. F. (2020). A network-based microfoundation of Granovetter’s threshold model for social tipping. Scientific reports, 10(1), 11202.
Citation: https://doi.org/10.5194/egusphere-2023-1622-AC5 -
AC6: 'Notes on organization of responses to reviewer comments', Keith Smith, 20 Mar 2024
First of all, we thank the reviewers for this constructive and well considered review of our manuscript. We will address these issues raised by each point below. In order to best organise our responses, we have structured these by Reviewer (underlined) and thematic concept being addresses [in brackets]. If similar issues are raised across reviewers, we address these more directly once, and cross-reference of our responses where appropriate for clarity.
Again, thank you all very much for these constructive and helpful feedback.
Citation: https://doi.org/10.5194/egusphere-2023-1622-AC6
-
RC2: 'Comment on egusphere-2023-1622', Anonymous Referee #2, 19 Nov 2023
The paper considers the urgent need for a rapid societal-scale transformation or social tipping in response to the threat of climate change impacts. It tests a hypothesis based on a ‘social activation model’ that there is a correlation between climate change concern and anticipation of long-term sea level rise. It employs three categories of tipping potential and finds the greatest potential for tipping in Western Pacific Rim and East Asian countries.
The topic of social tipping points is very popular and timely in the context of the upcoming UNFCCC COP 28 Conference and ongoing negotiations to limit global warming. The idea to combine estimate of future seal level rises with social survey data within a model of social activation is a novel methodological approach. The use of the MAGICC climate model, and the use of topographic and population distribution data, deserves credit. The paper is well structured and composed, with appropriate number and quality of references and strong supplementary materials, clear illustrations and correctly defined and used mathematical symbols.
Specific comments:
P.2. Figure 1: Why focus on projected global mean sea level rise in the five countries with the largest share of global greenhouse gas emissions, rather than, for example, countries most at risk in terms of human populations or economic impacts?
P.3. line 40-45: consider clarifying a ‘qualitative state’. The intention here is to define social tipping as change to a qualitatively different system state, This could be made clearer.
P.3. Line 46 change to ‘enabling conditions’ typo inverted first quotation mark.
P.3. line 48 suggest “originating from either natural or social systems”.
P.3. line 48-49: I am unclear about the distinction between natural and social system forces on the one hand, and “those emerging from individual-level changes”. Are individuals not part of social systems?
P.3. Line 50-53: The authors claim that “Within social-ecological systems, experienced climate impacts (Demski et al., 2017; Konisky et al., 2016), e.g. floods and heat waves (Ricke and Caldeira, 2014), have the potential to shift attitudes and behaviors toward climate change and instigate social tipping processes (Müller et al., 2021).” There is evidence to support this claim. However, it is by no means certain that people affected by climate-related extreme events will be motivated to seek bolder climate action. Depending on a range of variables – including the kind and degree of affect/trauma, in-group influences, and individual differences – some people may become demotivated or use denial as a defence mechanism (Marshall, 2014). There is also evidence that increased societal stress and insecurity resulting from climate-related extreme weather events might also reduce tolerance and cooperation and make pro-environmental policies even less likely (Friedman, 2005). Forces of climate delay are also well resourced and creative in exploiting fear and loss. The authors might therefore consider the possibility that climate impacts might have the opposite effect of triggering and accelerating ‘damaging cascades’ (Lenton et al., 2022, p.8) of social unrest, populism, conflict and ‘barbarization (Raskin, 2016, p. 26), leading to a ‘fractured world’ (Laybourn-Langton, 2022, p. 10).
P.4. line 89. This section introduces the concept of complex contagion. But the ‘social activation model’ is not introduced or explained until the authors state that ‘the complex contagion social activation model of social tipping’ is to be applied in the model design. What is ‘the social activation model’? Is it a term coined by the authors? The Norm Activation Model (Schwartz, 1977), the Value-Belief-Norm Theory (Stern, 2000), and the Theory of Planned Behaviour (Ajzen, 1991) all offer explanations for how values, norms, beliefs, and intentions can motivate behavioural change Do the authors consider any of these models to be relevant in this case?
P.4. Line 93. The authors state that individual preference factors and network structures can ‘trigger rapid shifts in social norms and behaviors’. p. 5. Line 99: similarly, the authors identify ‘ rapid adoption of environmental behaviors’ . Do the authors have any evidence that changes in attitudes of the populations in the target countries are likely to experience an emergent threshold function (F(r(t))) as opposed to more incremental change dynamics?
P.5. Line 106 – spelling ‘heterogeneities’
In the authors methodological approach, they adopt a ‘low-dimensional’ approach to deal with the ‘uncertainties and heterogeneities’ of real-world populations, using a tripartite categorisation based on resource mobilization theory that is 40-50 years old. There is a lot more to climate action beyond resource mobilization. Why was this categorisation considered the most appropriate? And how does it affect the generalisability of the results?
p.10. line 230: the authors estimate their ‘potentially active population share’ for each of the target countries according to national survey data of climate change concern. However, a self-reported level of climate change concern does not necessarily denote any connection to the subjects’ awareness of elevation above sea level or to the level of risk the subjects face under different SLR projections.
Citation: https://doi.org/10.5194/egusphere-2023-1622-RC2 -
AC2: 'Reply to Reviewer 2', Keith Smith, 20 Mar 2024
We thank the reviewer for the overall positive review of our paper. We have responded to each of these comments and suggestions below, organised by topic.
[Why focus on 5 exemplary countries in Figure 1?]
The reason for focussing on the five countries with the largest shares of GHG emission in the first part of the manuscript’s results section is twofold: First, we acknowledge that those countries are potentially also among the ones where effective climate policy might have the most substantial impact on global carbon emissions. At the same time, it is reasonable to assume that these countries do in fact possess the political capacities to lead others towards joining into collaborative efforts to mitigate climate change. Second, quantifying which countries are most at risk with respect to human population or socio-economic impacts is a highly multidimensional problem that is still subject of current research. Had we attempted to put our focus according to any of these dimensions our choice would have likely become very subjective. Ultimately, we note that previous works on the interrelation between future climate variability and policy have opted for similarly focusing on the five top emitters (Ricke & Caldeira, 2014). For the revision we propose to add a statement that justifies our choice of presented countries using the arguments above to the beginning of Sec. 4.1 in the manuscript.
[Spelling and technical issues]
We agree with the reviewer that the notion of a qualitative state in social systems deserves further clarification.We therefore propose to add examples from technology adoption, political systems, and environmental behaviour to the respective section in the introduction of the paper. We thank the reviewer for highlighting a few typographic errors, which we will fix during the revision.
We will change the sentence on line 48 to ‘The ultimate trigger of social tipping processes can come from forces originating from either natural or social systems [...]’.
[Individual and societal level changes]
Yes, we agree that individuals are part of social systems. For this analysis, we focus on a theory of change that originates (is triggered) from individual (micro-) level dynamics, as opposed to from systemic level changes (e.g. policy change). That is, we explore how these individual changes can enact system regime changes, which themselves reshape the individual actions nested within these regimes. This approach is situated directly within the “social tipping” theory of change elaborated within the paper, but should naturally be interpreted as complimentary, and part of, broader mechanisms towards how such rapidly shifts could occur. The interplay between micro-macro-micro changes are a foundation of social analyses looking at the co-evolution of structure and agency (e.g. “Coleman’s Boat”, Coleman 1994). In a revised version of the manuscript, we would make the connection between micro actions and resulting macro level changes more clear and linking to relevant literature on such theories of change.
[Uncertainty of changes resulting from climate impacts]
This is a very good point that we agree with. As we note in our response to Reviewer #1 above, there is much hetereogeniety in the effect of climate impacts on climate change attitudes and behaviour, and as this comment correctly notes, not all of these changes may be normatively ‘positive’. Furthermore, there is recent literature exploring the risks of societal tipping dynamics (e.g. Milkoreit 2023), which emphasizes the inherent danger of rapid social change, which is a very chaotic event and the outcomes are far from determined. Indeed, contemporary examples used to identify social tipping (e.g. Arab Spring) have not always resulted in long-term, “positive”, change within these societies. We suggest revising the discussion and theoretical background to emphasize the uncertainty not only in the effects of extreme events, but the direction of the resulting changes.
[Model terminology and relation to other theories of environmental social change]
This is a good point that is also raised in regard to the relation of the theory of change in this model to other forms of change that have been applied to environmental behaviours. The current version of the manuscript does not clearly elaborate the connections between social tipping and established theories of change - as it is more focused on applying these concepts in the modelling exercise as opposed to theoretically explaining them. But here, we would highlight the direct connections between classical threshold models of behavioural change (e.g. Granovetter 1978, Schelling 1972) alongside those focused on developing a micro-based foundation for social change (as noted above, and in line with the VBN theory noted here). This is not as directly connected to TBP based mechanisms, or other theoretical explanations of the attitude-behaviours gap (e.g. Kollmuss and Agyemann, 2002; Diekmann and Preisendörfer, 2003), as these are rather understood by the authorship team as potential barriers and limitations to the theory of change explored by this model. We suggest revising the manuscript to make the connections between social tipping literature and other proposed theories of environmental change (individual and structural) more clear.
We also suggest changing the name of the model to be “network-based threshold model for social tipping” to further clarify these connections to previous literature (Wiedermann et al., 2020) and reduce confusion over this modelling approach.
Lastly, as noted in our response to Reviewer 1 [Definition of active population], we would also suggest changing the title to reflect the new terminology of the model, as well as the proposed change in terminology for climate action: “A global threshold model of enabling conditions for social tipping in pro-climate change behaviours – the role of sea-level rise anticipation and climate change concern”
[Connections to resource mobilization theory]
We agree with the reviewer that resource mobilization theory has already been around for half a century. In our original paper (Wiedermann et al., 2020) we motivate our choice of the model by acknowledging that spreading of opinions and behaviours can often be explained through complex contagion (Watts, 2002; Centola et al., 2015 & 2018). At the same time, recent studies have investigated the role of ‘instigators’ (Singh et al., 2013) or ‘immune’ individuals (Karsai et al., 2016) in addition to the contingent population, as such groups are commonly discussed in the literature and seem apparent in real-world situations. Notably, such distinctions, even though they were only studied quantitatively in the recent decade, align very well with the attribution of groups in resource mobilisation theory (RMT). In this sense, we only draw the terminology from RMT without inferring any causal mechanism related to that theory. We think that using a tripartite categorisation is appropriate as it approximates relevant actors in any social movement in a meaningful way without the need for specifying more groups than necessary: those that will act, those that are willing to act and those that will never act.
We propose to rework our manuscript so that it becomes clear that all we draw from RMT is the terminology of groups while the rest follows from literature that is much more recent.
[Connection between climate change concern and SLR perceptions]
We completely agree with the reviewer that concern can not necessarily be linked to awareness or proper risk assessment. However, it seems that there is a misunderstanding at play that we will attempt to clarify in the revision of the manuscript.
The core idea of differentiating between potentially active and certainly active individuals is that the former don’t necessarily need to be directly affected by sea-level rise. Rather, it is the certainly active population, i.e., those that are or will be affected, that we assume to trigger a movement so that the potentially active individuals will join in, regardless their specific locale in a country. In other words, in order to instigate a movement, an individual needs to be both affected and concerned. Only then will those that are concerned, but not affected, join into a certain climate action based on social contagion and peer pressure.
We also agree that concern is not even distributed within countries. Indeed, there may even be more intra-country variance than across countries. But, there is the potential for spillovers within a country, for example, a person living in Berlin could be very concerned about SLR (but not directly affected). Yet, the person is connected to an individual in Hamburg who is concerned and affected, which could trigger both towards action. We suggest including this as a limitation and suggestion for future research in the revised manuscript.
Further, this comment notes that we use climate change concern, and not SLR risk perceptions, to operationalise the potentially active population. This is rather a product of the data limitations, where questions about the perceived risks associated with climate change are often asked more broadly, and less frequently in respect to specific impacts (such as SLR). While we agree that we do not directly capture SLR perceptions in the data used for this study, climate change concern can function as an appropriate “best available data” proxy. For example, recent research conducted by the authorship team suggests a close relationship between anticipation of SLR and climate change concern, where we find a strong, positive correlation r=0.76 in a recent sample of US adults (see Figure A1 in Smith et al., 2022). Accordingly, we would add this information as part of the revised limitations mentioned above.
Citation: https://doi.org/10.5194/egusphere-2023-1622-AC2 -
AC5: 'References for all responses', Keith Smith, 20 Mar 2024
As a final note, please do find our references for these responses to all reviewers here:
Akerlof, K., Covi, M., & Rohring, E. (2017). Communicating sea level rise. In Oxford Research Encyclopedia of Climate Science.
Akerlof, K., Merrill, J., Yusuf, J. E., Covi, M., & Rohring, E. (2019). Key beliefs and attitudes
for sea-level rise policy. Coastal Management, 47(4), 406-428.
Bolsen, T., Kingsland, J., & Palm, R. (2018). The impact of frames highlighting coastal flooding in the USA on climate change beliefs. Climatic Change, 147, 359-368.
Coleman, J. S. (1994). Foundations of social theory. Harvard university press.
Covi, M. P., & Kain, D. J. (2016). Sea-level rise risk communication: Public understanding, risk perception, and attitudes about information. Environmental Communication, 10(5), 612-633.
Diekmann, A., & Preisendörfer, P. (2003). Green and greenback: The behavioral effects of environmental attitudes in low-cost and high-cost situations. Rationality and Society, 15(4), 441-472.
Kollmuss, A., & Agyeman, J. (2002). Mind the gap: why do people act environmentally and what are the barriers to pro-environmental behavior?. Environmental education research, 8(3), 239-260.
Marzeion, B., & Levermann, A. (2014). Loss of cultural world heritage and currently inhabited places to sea-level rise. Environmental Research Letters, 9(3), 034001.
Milkoreit, M. (2023). Social tipping points everywhere?—Patterns and risks of overuse. Wiley Interdisciplinary Reviews: Climate Change, 14(2), e813.
Müller, P. M., Heitzig, J., Kurths, J., Lüdge, K., & Wiedermann, M. (2021). Anticipation-induced social tipping: can the environment be stabilised by social dynamics?. The European Physical Journal Special Topics, 230(16-17), 3189-3199.
Priestley, R. K., Heine, Z., & Milfont, T. L. (2021). Public understanding of climate change-related sea-level rise. PLoS One, 16(7), e0254348.
Quoß, F., & Rudolph, L. (2022). Operationalisation matters: Weather extremes as noisy natural experiment show no influence on political attitudes. https://osf.io/preprints/osf/m5rz8
Smith, E. K., Eder, C., Donges, J. F., Heitzig, J., Katsanidou, A., Wiedermann, M., & Winkelmann, R. (2022). Domino Effects in the Earth System-The role of wanted social tipping points. OSF Preprints.
Smith, S. R. (2023). Enabling a political tipping point for rapid decarbonisation in the United Kingdom. EGUsphere, 2023, 1-21.
Tàbara, J. D., Frantzeskaki, N., Hölscher, K., Pedde, S., Kok, K., Lamperti, F., ... & Berry, P. (2018). Positive tipping points in a rapidly warming world. Current Opinion in Environmental Sustainability, 31, 120-129.
Tàbara, J. D. (2023). Regenerative sustainability. A relational model of possibilities for the emergence of positive tipping points. Environmental Sociology, 9(4), 366-385.
Thomas, M., Pidgeon, N., Whitmarsh, L., & Ballinger, R. (2015). Mental models of sea-level change: A mixed methods analysis on the Severn Estuary, UK. Global Environmental Change, 33, 71-82.
Wiedermann, M., Smith, E. K., Heitzig, J., & Donges, J. F. (2020). A network-based microfoundation of Granovetter’s threshold model for social tipping. Scientific reports, 10(1), 11202.
Citation: https://doi.org/10.5194/egusphere-2023-1622-AC5 -
AC6: 'Notes on organization of responses to reviewer comments', Keith Smith, 20 Mar 2024
First of all, we thank the reviewers for this constructive and well considered review of our manuscript. We will address these issues raised by each point below. In order to best organise our responses, we have structured these by Reviewer (underlined) and thematic concept being addresses [in brackets]. If similar issues are raised across reviewers, we address these more directly once, and cross-reference of our responses where appropriate for clarity.
Again, thank you all very much for these constructive and helpful feedback.
Citation: https://doi.org/10.5194/egusphere-2023-1622-AC6
-
AC2: 'Reply to Reviewer 2', Keith Smith, 20 Mar 2024
-
RC3: 'Comment on egusphere-2023-1622', Anonymous Referee #3, 24 Nov 2023
I really enjoyed reading the paper, it makes really a great and very much needed contribution to understanding mechanisms of positive social tipping. I also see that two reviews have already been published, so I'm not sure my review is even needed any longer and I will keep it brief.
In substantial terms my only main objection would be that the focus is on SLR, but climate change has a whole range of impacts (draughts, flooding, etc.) that will also affect populations that are land-locked or live more land inwards, this shortcoming becomes particularly apparent in the Tipping Class II. So maybe the authors could emphasise that this type of analyses should be expanded in future considering a range of climate changer impacts.
I also thought that the results on modular networks are quite interesting and worth to elaborate on. I think the suggestion that in the case of anticipation of SRL such modularised networks are less likely can be challenged. For instance, we know that there are inequalities (e.g. racial, class) at play when it comes to who is most likely to be affected by climate change impacts (incl. SLR) and in these circumstances a modularized network may actually be quite realistic and this result would then show how inequality could prevent tipping potentially?
Apart from that, I would urge the authors to go through the manuscript and check terminology and notation for consistency and clarity. For instance the reader may get confused between the notion of contingently active and potential active, as they signify very different population groups in the modelling, but this distinction is not clear enough in the chosen terminology.
I would also urge the authors avoiding using notations multiple times for different parameters etc. as this is extremely confusing. For instance s is used in equation 4 to represent a given survey in later equations it is used for simulation it seems and then in equation 9 for set of all ensemble members? u is used for upper branch but also in u_tot for tipping potential.
The explanation for in Figure 5 caption needs to come earlier, namely on p.11 (line 255).
In 3.5 state that tot stands for total and explain u_bif. The explanations are given only properly in Figure 5 caption.
Figure 5. You used United States in Figure 4, here you write United States of America, make sure this consistent.
Figure 6. I would suggest that the colours should be equivalent across the three images, otherwise it becomes very misleading as red may mean very different things in Figure A, B or C etc. Also add to B that this is total tipping potential.
Correct on p. 11 line 246, it should be Tab A2 and Tab A3.
Overall however, an excellent paper!
Citation: https://doi.org/10.5194/egusphere-2023-1622-RC3 -
AC3: 'Reply to Reviewer 3', Keith Smith, 20 Mar 2024
Thank you as well for these constructive comments, which we respond to by each topic raised below.
[SLR as chosen impact and distributional issues]
Thank you very much for the supportive comments. We very much agree with the potential to expand this analysis to other climate impacts. This is something we will further emphasise in the revised manuscript. In our response to the other expert reviews, we also note how we will further elaborate the rationale for why we focus on SLR.
We also agree with the need to focus on inequalities associated with these impacts. While the model does not allow for the specificity to talk about specific societal subgroups within a given country (e.g. race, class) - we can make a clear link between the likelihood of people to be exposed to SLR (and related impacts) and the comparative diminished capacity/agency of these people to respond. This would speak directly to the concern raised, that the people most likely to ‘tip’ are less likely to have the capacity to act. This is something that we view as important to highlight, and will expand upon in the revised discussion.
[Terminology and Specific Technical Changes]
We thank you for these more specific comments as well. We will address each of these accordingly within the revised manuscript.
Citation: https://doi.org/10.5194/egusphere-2023-1622-AC3 -
AC5: 'References for all responses', Keith Smith, 20 Mar 2024
As a final note, please do find our references for these responses to all reviewers here:
Akerlof, K., Covi, M., & Rohring, E. (2017). Communicating sea level rise. In Oxford Research Encyclopedia of Climate Science.
Akerlof, K., Merrill, J., Yusuf, J. E., Covi, M., & Rohring, E. (2019). Key beliefs and attitudes
for sea-level rise policy. Coastal Management, 47(4), 406-428.
Bolsen, T., Kingsland, J., & Palm, R. (2018). The impact of frames highlighting coastal flooding in the USA on climate change beliefs. Climatic Change, 147, 359-368.
Coleman, J. S. (1994). Foundations of social theory. Harvard university press.
Covi, M. P., & Kain, D. J. (2016). Sea-level rise risk communication: Public understanding, risk perception, and attitudes about information. Environmental Communication, 10(5), 612-633.
Diekmann, A., & Preisendörfer, P. (2003). Green and greenback: The behavioral effects of environmental attitudes in low-cost and high-cost situations. Rationality and Society, 15(4), 441-472.
Kollmuss, A., & Agyeman, J. (2002). Mind the gap: why do people act environmentally and what are the barriers to pro-environmental behavior?. Environmental education research, 8(3), 239-260.
Marzeion, B., & Levermann, A. (2014). Loss of cultural world heritage and currently inhabited places to sea-level rise. Environmental Research Letters, 9(3), 034001.
Milkoreit, M. (2023). Social tipping points everywhere?—Patterns and risks of overuse. Wiley Interdisciplinary Reviews: Climate Change, 14(2), e813.
Müller, P. M., Heitzig, J., Kurths, J., Lüdge, K., & Wiedermann, M. (2021). Anticipation-induced social tipping: can the environment be stabilised by social dynamics?. The European Physical Journal Special Topics, 230(16-17), 3189-3199.
Priestley, R. K., Heine, Z., & Milfont, T. L. (2021). Public understanding of climate change-related sea-level rise. PLoS One, 16(7), e0254348.
Quoß, F., & Rudolph, L. (2022). Operationalisation matters: Weather extremes as noisy natural experiment show no influence on political attitudes. https://osf.io/preprints/osf/m5rz8
Smith, E. K., Eder, C., Donges, J. F., Heitzig, J., Katsanidou, A., Wiedermann, M., & Winkelmann, R. (2022). Domino Effects in the Earth System-The role of wanted social tipping points. OSF Preprints.
Smith, S. R. (2023). Enabling a political tipping point for rapid decarbonisation in the United Kingdom. EGUsphere, 2023, 1-21.
Tàbara, J. D., Frantzeskaki, N., Hölscher, K., Pedde, S., Kok, K., Lamperti, F., ... & Berry, P. (2018). Positive tipping points in a rapidly warming world. Current Opinion in Environmental Sustainability, 31, 120-129.
Tàbara, J. D. (2023). Regenerative sustainability. A relational model of possibilities for the emergence of positive tipping points. Environmental Sociology, 9(4), 366-385.
Thomas, M., Pidgeon, N., Whitmarsh, L., & Ballinger, R. (2015). Mental models of sea-level change: A mixed methods analysis on the Severn Estuary, UK. Global Environmental Change, 33, 71-82.
Wiedermann, M., Smith, E. K., Heitzig, J., & Donges, J. F. (2020). A network-based microfoundation of Granovetter’s threshold model for social tipping. Scientific reports, 10(1), 11202.
Citation: https://doi.org/10.5194/egusphere-2023-1622-AC5 -
AC6: 'Notes on organization of responses to reviewer comments', Keith Smith, 20 Mar 2024
First of all, we thank the reviewers for this constructive and well considered review of our manuscript. We will address these issues raised by each point below. In order to best organise our responses, we have structured these by Reviewer (underlined) and thematic concept being addresses [in brackets]. If similar issues are raised across reviewers, we address these more directly once, and cross-reference of our responses where appropriate for clarity.
Again, thank you all very much for these constructive and helpful feedback.
Citation: https://doi.org/10.5194/egusphere-2023-1622-AC6
-
AC3: 'Reply to Reviewer 3', Keith Smith, 20 Mar 2024
-
RC4: 'Comment on egusphere-2023-1622', Anonymous Referee #4, 05 Feb 2024
This is a nice paper in that it examines a potential mechanism by which positive social tipping might come about – specifically, using a social activation model to explain increased concern as a function of (anticipation of) sea level rise. I had a few questions / suggestions that might strengthen the paper or bring it in line with some existing literature.
First, I wondered why the authors focused on SLR rather than extreme weather events that are happening now/on a much shorter timescale, such as hurricanes, wildfires, or droughts? Much of the literature on concern as a result of experience with climate change is focused on extreme weather events, rather than SLR, which is a much slower and distant process (with long anticipation horizons) – due to this prolonged time horizon, people may also expect that they/societies will be able to adapt. A focus on extreme weather impact would link the paper to this literature and provide a clearer basis for the increased concern the authors suppose.
Second, I was not familiar with the social activation model. However, the pathway from concern to active population or action was not sufficiently motivated. Concern is, on its own, unlikely to lead to action – many people globally are currently concerned about climate change and there is still insufficient action, and concern need not be a precursor for action (i.e. action may be a beneficial proposition in its own right). Much of the literature on experience of extreme weather and attitudes and behaviors finds tenuous relationship with behavior, and a much more robust relationship to concern. Even in areas with regular flooding and extreme weather events (e.g. Florida, Texas Gulf Coast), there is limited concern about climate change. The authors should be clear about this limitation – focus on concern rather than actions/behavior - up front. They could frame the paper in terms of a tipping point in concern rather than action (as it is currently unclear what the tipped state/actions are), which would require a more nuanced explanation of how one begets the other (which will vary across context, depend on government action, etc.). In the discussion, they could then speculate that growing concern could lead to subsequent action through pressure on politicians etc. Currently there is some discussion in the introduction, but it then seems lost throughout. Also, action is a broad term – it is unclear which actions the authors are referring to… is this mitigative actions? Adaptation? Finally, responses are likely to vary substantially across countries due to differences in income/adaptive capacity/culture etc. These may also very well be correlated with SLR exposure. It would be great if the authors could account for some of this heterogeneity.
The authors argue that for the case of SLR anticipation, real-life social networks are unlikely to have modularized network structures, but I am not sure why this is the case. For example, there are very different demographics / types of jobs etc. in rural vs coastal areas. There is literature showing that people sort in where they live – often choosing to leave nearby close others (and migrating to areas where they know people). In the US, many exposed areas also have climate skeptics / deniers / and Republicans, who are less likely to endorse climate action. Relatedly, the authors do not discuss the literature on motivated reasoning but there is some literature showing that people often interpret extreme weather events in ways that fit their prior beliefs – these could be the never active types, but they are likely also some of the potentially active types. In other words, the same amount of information might result in heterogeneous responses depending on prior beliefs even among potentially active types.
It was not clear to me why the authors exclude low elevation regions. Do they expect concern to propagate in a different way in these areas?
The discussion of sensitive interventions, with mentions of policy regime changes, social movements, policy entrepreneurs etc., comes a bit out of the blue here and seems non-specific / lacking nuance relative to the rest of the piece. These sections seem a little tacked on, and the mechanisms aren’t fleshed out so they seem a bit unsatisfying.
Citation: https://doi.org/10.5194/egusphere-2023-1622-RC4 -
AC4: 'Reply to Reviewer 4', Keith Smith, 20 Mar 2024
Lastly, we greatly appreciate the feedback provided as well. Similar to the other comments, we have responded by topic below, as well with linkages to other responses to similar comments we have addressed above.
[SLR as Chosen Climate Impact]
Thank you very much for this comment. This is something that is shared commonly across many of the reviewers. For the sake of parsimony, we will not repeat our responses here, but would rather refer to our replies to Reviewer #1 [Why does the paper focus on SLR and not on other impacts? Comment #1] and [Why does the paper focus on SLR and not on other impacts? Comment #2] and to our response to Reviewer #2 [Uncertainty of changes resulting from climate impacts].
[Heterogeneity in the effect of concern]
Thank you for raising this point, we very much agree with this. We also note some of these issues about the variance in the effect of concern in shaping climate actions in our response to Reviewer #1 [Salience of SLR and how SLR perceptions shape attitudes and behaviors] and Reviewer #2 [Connection between climate change concern and SLR perceptions] (amongst other comments above). We very much appreciate the further suggestion to expand the discussion on what this means in terms of political relevance. If a system has higher levels of concern, we find that it has the necessary conditions for change, and in this case, other efforts (such as pressuring politicians) could be successful to push towards a tipped state. But of course, what action can be taken to “trigger” the social tipping depends on the local context. We suggest expanding the discussion in the revised manuscript to make these recommendations and relevance more clear.
[Defining Action]
Thank you for this comment as well, this was similarly shared by Reviewer #1 [Definition of active population]. We will refine this definition in the revised manuscript in line with our response to Reviewer #1, but in a particular response to this comment, we would consider adaptation to SLR to likely not be included as action, as in this context, this likely means that the impact of SLR has already occurred, and is likely unchangeable during the normal lifetime of humans. Rather, we are interested in how people can be motivated to act to minimise the risks of being exposed to SLR - so focusing on mitigating actions.
[Heterogeneity across subgroups within population]
We also agree with some of the concerns raised here about how the mechanism of never active->potentially active->certainly active may vary for certain subgroups within a population. Given the model resolution, such intra-country dynamics are not included in these analyses. Certainly, these could work both ways - where some countries will have sub-groups that are more likely to act and some that are not. We agree that this is a limitation that would be discussed in a future revision of the paper, as well as a call for further research into within country dynamics, which could build upon this assessment.
[Exclusion of lower elevation regions]
We had excluded lower lying regions (e.g. Azerbijian, Kazakhstan and the Netherlands) as well as those that are situated inland (e.g. Mongolia, Austria) as these present more ‘extreme’ case that are likely far outside of data distribution of the remaining sample. Particularly in the case of lower-lying countries, we consider these to be more specific cases, that are not likely as well explained by the model. This is an approach that has also been adopted by similar modelling exercises (e.g. Marzeion and Levermann, 2014). We would suggest revising the manuscript to make this justification more clear, and if the reviewer believes appropriate, present additional analyses including the data from lower lying regions as a robustness check.
[Connection to theories of change]
Thank you for the final comment on the discussion of sensitive interventions and other theories of change. This is similar to a comment raised by Reviewer #1 [Model terminology and relation to other theories of environmental social change]. Accordingly, we suggest revising this section in the discussion, to be more clear how social tipping relates to other theories of change, and how the proposed change would occur (bottom-up). We believe this would make these statements more clear to the reader.
Citation: https://doi.org/10.5194/egusphere-2023-1622-AC4 -
AC5: 'References for all responses', Keith Smith, 20 Mar 2024
As a final note, please do find our references for these responses to all reviewers here:
Akerlof, K., Covi, M., & Rohring, E. (2017). Communicating sea level rise. In Oxford Research Encyclopedia of Climate Science.
Akerlof, K., Merrill, J., Yusuf, J. E., Covi, M., & Rohring, E. (2019). Key beliefs and attitudes
for sea-level rise policy. Coastal Management, 47(4), 406-428.
Bolsen, T., Kingsland, J., & Palm, R. (2018). The impact of frames highlighting coastal flooding in the USA on climate change beliefs. Climatic Change, 147, 359-368.
Coleman, J. S. (1994). Foundations of social theory. Harvard university press.
Covi, M. P., & Kain, D. J. (2016). Sea-level rise risk communication: Public understanding, risk perception, and attitudes about information. Environmental Communication, 10(5), 612-633.
Diekmann, A., & Preisendörfer, P. (2003). Green and greenback: The behavioral effects of environmental attitudes in low-cost and high-cost situations. Rationality and Society, 15(4), 441-472.
Kollmuss, A., & Agyeman, J. (2002). Mind the gap: why do people act environmentally and what are the barriers to pro-environmental behavior?. Environmental education research, 8(3), 239-260.
Marzeion, B., & Levermann, A. (2014). Loss of cultural world heritage and currently inhabited places to sea-level rise. Environmental Research Letters, 9(3), 034001.
Milkoreit, M. (2023). Social tipping points everywhere?—Patterns and risks of overuse. Wiley Interdisciplinary Reviews: Climate Change, 14(2), e813.
Müller, P. M., Heitzig, J., Kurths, J., Lüdge, K., & Wiedermann, M. (2021). Anticipation-induced social tipping: can the environment be stabilised by social dynamics?. The European Physical Journal Special Topics, 230(16-17), 3189-3199.
Priestley, R. K., Heine, Z., & Milfont, T. L. (2021). Public understanding of climate change-related sea-level rise. PLoS One, 16(7), e0254348.
Quoß, F., & Rudolph, L. (2022). Operationalisation matters: Weather extremes as noisy natural experiment show no influence on political attitudes. https://osf.io/preprints/osf/m5rz8
Smith, E. K., Eder, C., Donges, J. F., Heitzig, J., Katsanidou, A., Wiedermann, M., & Winkelmann, R. (2022). Domino Effects in the Earth System-The role of wanted social tipping points. OSF Preprints.
Smith, S. R. (2023). Enabling a political tipping point for rapid decarbonisation in the United Kingdom. EGUsphere, 2023, 1-21.
Tàbara, J. D., Frantzeskaki, N., Hölscher, K., Pedde, S., Kok, K., Lamperti, F., ... & Berry, P. (2018). Positive tipping points in a rapidly warming world. Current Opinion in Environmental Sustainability, 31, 120-129.
Tàbara, J. D. (2023). Regenerative sustainability. A relational model of possibilities for the emergence of positive tipping points. Environmental Sociology, 9(4), 366-385.
Thomas, M., Pidgeon, N., Whitmarsh, L., & Ballinger, R. (2015). Mental models of sea-level change: A mixed methods analysis on the Severn Estuary, UK. Global Environmental Change, 33, 71-82.
Wiedermann, M., Smith, E. K., Heitzig, J., & Donges, J. F. (2020). A network-based microfoundation of Granovetter’s threshold model for social tipping. Scientific reports, 10(1), 11202.
Citation: https://doi.org/10.5194/egusphere-2023-1622-AC5 -
AC6: 'Notes on organization of responses to reviewer comments', Keith Smith, 20 Mar 2024
First of all, we thank the reviewers for this constructive and well considered review of our manuscript. We will address these issues raised by each point below. In order to best organise our responses, we have structured these by Reviewer (underlined) and thematic concept being addresses [in brackets]. If similar issues are raised across reviewers, we address these more directly once, and cross-reference of our responses where appropriate for clarity.
Again, thank you all very much for these constructive and helpful feedback.
Citation: https://doi.org/10.5194/egusphere-2023-1622-AC6
-
AC4: 'Reply to Reviewer 4', Keith Smith, 20 Mar 2024
Viewed
| HTML | XML | Total | BibTeX | EndNote | |
|---|---|---|---|---|---|
| 708 | 210 | 83 | 1,001 | 52 | 47 |
- HTML: 708
- PDF: 210
- XML: 83
- Total: 1,001
- BibTeX: 52
- EndNote: 47
Viewed (geographical distribution)
| Country | # | Views | % |
|---|
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1